SUPERIOR COURT OF THE DISTRICT OF COLUMBIA CRIMINAL DIVISION - FELONY BRANCH UNITED STATES OF AMERICA, v. RAYMOND ANTHONY JENKINS. Criminal No. F-320-00 Judge Broderick Trial Date 10/12/04 _______________________________/ GOVERNMENT'S SUPERSEDING OPPOSITION TO DEFENDANT’S MOTIONS IN LIMINE TO PRECLUDE INTRODUCTION OF DNA EVIDENCE KENNETH L. WAINSTEIN, United States Attorney. MICHAEL AMBROSINO, D.C. BAR #451549 VALINDA JONES, D.C. BAR #398464, Assistant United States Attorneys. 555 Fourth Street, N.W. Room 9816 Washington, D.C. 20001 (202) 514-0504 INDEX Page INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 BACKGROUND OF THE CASE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 DNA EVIDENCE TO BE OFFERED AT TRIAL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 ARGUMENT I. Defendant’s Supplemental Motion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 II. Defendant’s Cold-Hit Motion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 III. The law in the District of Columbia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 IV. DNA analysis by the PCR/STR method is universally accepted in the relevant scientific community and the courts V. A. Background of DNA analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 B. The PCR process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 C. The current generation of PCR-based analysis: STRs 1. Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 2. The STR method is generally accepted. . . . . . . . . . . . . . . . .30 The defendant’s arguments against the DNA typing methods in this case go to weight of the evidence, not its admissibility. . . . . . . . . . .37 A. The Court does not have to find that the tools used to apply the PCR/STR technique are generally accepted; even if it did, the kits and instrument used in these cases are generally accepted. . . . . . . . . . . . . . . . . . . . . . . . . . . . .38 1. 2. No inquiry is required under Frye. . . . . . . . . . . . . . . . . . . . . . 38 The kits are generally accepted. . . . . . . . . . . . . . . . . . . . . . .41 Adherence to the TWGDAM guidelines is not required for general acceptance. . . . . . . . . . . 42 b. The Profiler Plus/Cofiler/310 Analyzer system has been proven reliable by laboratory validation studies, published peer-reviewed research, and actual experience. . . . . . . . . . . . . . . .46 B. The general acceptance of DNA typing methods does not turn on whether the laboratory applying those methods is subject to blind or open proficiency testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 C. The general acceptance of a scientific method does not turn on whether the quality assurance standards and protocols for that method are also generally accepted, or whether the laboratory applying that method has followed accepted protocols. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56 Allegations of interpretive errors in individual cases go to weight, not admissibility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61 D. VI. a. 1. The Frye test does not require proof of adherence to correct procedures in each particular case. . . . . . . . . . . .61 2. The defendant’s allegation that the methodology applied in this particular case is unreliable does not demonstrate deficiencies in the underlying scientific methods. . . . . . . . . . . . . . . . . . . . .64 The procedures by which the FBI determines statistical match probabilities, which include use of the product rule, are generally accepted by the relevant scientific community. . . . . . .67 A. Although not admissible twelve years ago, probability calculations based on the product rule are now admissible because the scientific consensus has changed. . . . . . . . . . . . . . . .70 B. The FBI calculates probability statistics for single-source evidence samples by methods found acceptable by the National Research Council. . . . . . . . . . . . . . . . 77 The FBI’s method for calculating combined probabilities for mixed sample DNA profiles is generally accepted. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 C. D. VII. Inclusion of error rates in probability statistics is unacceptable in the relevant scientific community, and therefore not required under Frye. . . . . . . . . . . . . . . . . . . . . . . . . .91 The fact that the defendant was first identified by a “cold-hit” on a DNA database does not affect the admissibility of the DNA evidence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .94 A. B. The Random Match Probability (RMP) is always relevant and admissible to show the rarity of a DNA profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 The approaches cited by defendant are generally accepted methods of answering questions other than the rarity of a DNA profile, but not all are admissible under legal evidentiary standards. . . . . . . . . . . . . . . . . . . . . . . . . 103 1. The NRC I Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 2. 3. The NRC II Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112 The Donnelly and Friedman Approach. . . . . . . . . . . . . . . .116 iv C. The failure to incorporate estimates of error rates into the probability statistics does not preclude admission of the DNA evidence in a cold hit case. . . . . . . . . . . . . . . . . . . . D. The probative value of the DNA evidence in a cold-hit case is not outweighed by prejudicial effect. . . . . . . .121 E. The chain of custody for defendant’s DNA sample in the Virginia offender database is irrelevant to this case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 122 VIII. The Court does not need to hold an evidentiary hearing on all of the defendant’s claims. . . . . . . . . . . . . . . . . . . . . . . . 123 v SUPERIOR COURT OF THE DISTRICT OF COLUMBIA CRIMINAL DIVISION - FELONY BRANCH UNITED STATES OF AMERICA Criminal No. F-320-00 v. Judge Broderick Trial Date: 10-12-04 RAYMOND ANTHONY JENKINS. / GOVERNMENT'S SUPERCEDING OPPOSITION TO DEFENDANT’S MOTIONS IN LIMINE TO PRECLUDE INTRODUCTION OF DNA EVIDENCE INTRODUCTION On March 12, 2001, defendant filed a 112 page motion in limine to preclude the introduction of DNA evidence (hereinafter “Defendant’s Original Motion”). The United States filed a 91 page opposition to that motion on June 23, 2001. Almost three years later, and on the eve of the last trial date, defendant filed two additional DNA motions entitled “Supplemental Motion To Exclude Nuclear DNA ‘Match’ Evidence, DNA Expert Testimony, and ‘Random Match Probability’ Calculation” (hereinafter “Defendant’s Supplemental Motion”) and “Motion To Exclude DNA ‘Inclusion’ Evidence, Expert Testimony, And Frequency Statistics Because There Is No General Acceptance As To The Statistical methods Of Interpretation Of DNA Evidence That Derives From A ‘Cold Hit’ In A DNA Database, With Points And Authorities In Support” (hereinafter “Defendant’s Cold-Hit Motion”). Defendant’s challenges to the DNA evidence can be divided into two general categories: (1) the method by which the FBI amplifies, analyzes, and interprets DNA profiles; and (2) the method by which the FBI computes the statistical significance of DNA matches. The Defendant’s Original Motion discusses both of these issues. The Defendant’s Supplemental Motion merely repeats the same issues set forth in Defendant’s Original Motion, all of which have been resolved in favor of the government by the Honorable Robert I. Richter (in United States v. Orlando Roberts, F-771-01 and United States v. David Veney, F-3986-00)1/, the Honorable Ann O’Reagan Keary (in United States v. Henry Thompson, F-6540-02)2/, and the Honorable Judith E. Retchin (in United States v. Sidney Smith), F-9156-98.3/ Defendant’s Cold-Hit Motion purports to raise a new statistical issue resulting from the fact that the defendant was first identified through a search of a DNA offender database. As discussed at length below, the defendant’s motions are without merit. Due to the passage of time, and in the interests of efficiency, the United States, by and through its attorney, the United States Attorney for the District of Columbia, hereby files this 1/ Judge Richter’s ruling is attached hereto as Appendix II, Tab X. 2/ Judge Keary’s oral ruling of June 17, 2003, is attached hereto as Appendix II, Tab Y. 3/ Judge Retchin’s oral ruling of October 2001, is attached hereto as Appendix I, Tab S. In each of these cases, the judges resolved the issues without a hearing. In the Orlando Roberts/David Veney and Henry Thompson cases, the defense specifically requested a Frye hearing. In the Sidney Smith case, the parties adopted the original DNA pleadings in this case and agreed to have the court resolve the DNA admissibility issue based upon the Jenkins 2 superceding pleading along with superceding appendices respectfully opposing all three of defendant’s DNA pleadings. As grounds for this superceding opposition, the United States relies on the following points and authorities, and such other points and authorities as may be adduced at any hearing on this matter: BACKGROUND OF THE CASE The government expects the evidence to show that, on Friday June 4, 1999, at approximately 4:15 p.m., Metropolitan Police Department (“MPD”) officers were called to the home of Dennis Dolinger at 1516 Potomac Avenue, S.E., Washington, D.C., where they discovered Mr. Dolinger’s body in the basement. He had been stabbed many times in the head. The District’s Medical Examiner ruled the death a homicide. Based on the condition of the crime scene, it appeared that an assailant had been injured during the assault. Blood stains were found on clothing and other surfaces in the basement where the decedent’s body was discovered, there were traces of blood leading from the basement to the first and second floors of the house, and outside to the front walkway and sidewalk, and bloody clothing was found in a room on the second floor. The police also determined pleadings. 3 that, among other things, a diamond ring, a gold chain, and possibly some cash were missing from the decedent’s house. The MPD soon learned that a man identified as Stephen Watson made several purchases using the decedent’s credit card within 24 hours of the murder. After further investigation revealed that Watson had other items belonging to the decedent in his possession, and that he had been seen in the vicinity of decedent’s house on the day of the murder, Watson was arrested on a warrant for Felony Murder While Armed. MPD detectives sent evidence samples taken from the murder scene (“evidence samples”) to the Federal Bureau of Investigation (“FBI”) for forensic analysis. After conducting an analysis of DNA extracted from blood found in several of the evidence samples and DNA from blood samples taken from Watson, the decedent, and other individuals (“known samples”), the FBI forensic laboratory reported on November 8, 1999, that the blood of the decedent was contained in a number of the evidence samples, that Stephen Watson could be excluded as the contributor of the blood from any of the tested crime scene samples, and that a single individual, whose identity was unknown at that time, appeared to be the contributor of the blood in a significant number of the evidence samples. On November 16, 1999, the Virginia Division of Forensic Science ("the Division") was asked to assist in the investigation. The Division was sent a copy of 4 the FBI’s report, which contained the DNA profile of the unknown contributor of the blood evidence from the crime scene. Using a portion of this profile4/, the Division searched the Virginia DNA Databank and concluded that the profile was consistent with the DNA profile of one “Robert P. Garrett.” Law enforcement records revealed that the name “Robert P. Garrett” was an alias used by the defendant Raymond Anthony Jenkins, and that Jenkins’s personal information matched the personal information of the individual known to Virginia authorities as Robert P. Garrett. As a result, MPD obtained a search warrant for a sample of defendant’s blood, which it acquired on November 23, 1999. The FBI performed a DNA analysis of defendant’s known blood sample using the same methods and tools it had used to analyze the DNA in the samples from the crime scene. A comparison of defendant’s DNA profile with the DNA profile of evidence samples revealed that defendant’s DNA profile matched the profiles of blood from several locations at the Dolinger crime scene. In particular, the FBI identified defendant’s blood on: (a) a pair of jeans found in the basement near 4/ As discussed in greater detail infra, the DNA profiles were developed by the FBI by typing information at certain locations of the DNA known as “loci.” The DNA profile developed by the FBI of the unknown contributor consisted of a 13 locus profile. However, at the time the database search was performed back in 1999, the Virginia Division of Forensic Science catalogued offenders in its database system utilizing an 8-locus DNA profile. Thus, in searching the Virginia database, the Virginia Division of Forensic Science only utilized a portion of the profile developed by the FBI, i.e., 8 out of the 13 loci. 5 the decedent; (b) a shirt found in the upstairs exercise room; (c) a towel on the basement bathroom rack; (d) the sink stopper in the sink of the same bathroom; and (e) a railing between the first and second floors of the residence. The FBI had previously determined that the decedent’s blood was also on the jeans and shirt that contained defendant’s blood. After identifying defendant as a suspect, the police also received information from a witness who reported that, the day after the murder, the witness saw the defendant in possession of a diamond ring, gold chains, and over $1,000 in cash. The defendant also appeared to have numerous scratches or cuts to his face. Defendant was arrested on January 13, 2000, and charged with firstdegree murder. Throughout the course of these proceedings, the government furnished defendant’s counsel with discovery materials relevant to the DNA profile evidence in this case. These materials included, among other items, MPD crime-scene examination reports, MPD evidence reports, correspondence with the Virginia Department of Criminal Justice Services regarding that office’s initial identification of defendant through the Commonwealth’s DNA profile database, reports of the FBI tests of the blood-sample evidence, FBI DNA laboratory protocols for performance of the DNA tests in this case, and the curriculum vitae for the FBI scientist who oversaw the testing. In fact, in an effort to accommodate the defense, the United States has gone well beyond the scope 6 of Rule 16 by allowing the defendant’s counsel to conduct full interviews of FBI employees, including the DNA analyst, the hair and fiber analyst, and the fingerprint analyst.5/ DNA EVIDENCE TO BE OFFERED AT TRIAL 5/ Defendant’s assertion that it may be necessary to supplement the various DNA pleadings with additional information because discovery in this case is allegedly incomplete (Defendant’s Supplemental Motion at 23) is utterly groundless. The United States has fulfilled its obligations under Rule 16. Moreover, as noted above, the Unite States has gone beyond its obligations under Rule 16 by permitting defendant’s attorneys to interview a number of forensic witnesses. Hence, the Court should not permit any further eleventh hour pleadings by the defense. 7 The FBI laboratory identified and compared the DNA found in the crime scene evidence samples with the DNA found in defendant’s and decedent’s blood samples using two complementary scientific methods: first, samples of DNA were amplified, or copied at specifically-targeted locations, using a technique called polymerase chain reaction (“PCR”); and, second, certain segments of DNA (called short tandem repeats, or “STRs”) were identified, or “typed,” compared to each other, and found to match either the defendant or the decedent.6/ To facilitate its lab analysis, the FBI used two commercial kits manufactured by the Perkin Elmer Corporation (“Perkin Elmer”),7/ called AmpFLSTR Profiler Plus (“Profiler Plus”) and AmpFLSTR Cofiler (“Cofiler”). These kits enable FBI DNA analysts to locate, mark, and amplify a total of 13 separate STR sites on the DNA samples under study. The analysts then used a machine called a 310 Prism Genetic Analyzer (“310 Analyzer”) to separate the STRs found at each of the targeted 13 sites by the method of laser fluorescent capillary electrophoresis. Each of the 13 STR sites found in each sample of DNA was identified, or typed, according to the number of STRs found at that site. Together, the types of STRs found at all 13 sites constitute the DNA profile for each sample of DNA. After finding that defendant’s or decedent’s DNA profile 6/ These methods will be described in greater detail infra. 7/ The company is now called Applied Biosystems. Because most of the cases addressing the company’s kits use the company’s former name, we do the same in this brief. 8 was identical to the DNA profile from the evidence samples, the FBI calculated the statistical probability that the same profile might be found at random in a population of unrelated individuals. These calculations were made using statistical methods that included a procedure known as the product rule. The United States intends to offer expert testimony at trial pertaining to items of evidence that were the subject of the above analysis. There are two types of DNA evidence that were recovered from the crime scene: (1) “sole source samples” that comprised blood from a single individual; and (2) “mixture samples” that contained blood from two or more individuals. With respect to the sole source samples, the government’s DNA expert will testify that the DNA profile of blood developed from thirteen sole source evidence samples is identical to the defendant’s DNA profile, and that the DNA profile of blood developed from twenty two sole source evidence samples is identical to the decedent’s DNA profile. Hence, there are a total of thirty five sole source samples. In all but two of these sole source samples, the FBI was able to amplify the DNA profile at 13 loci. In two of the sole source samples (matching the defendant’s DNA profile), the FBI was able to amplify the samples at only 8 loci. The statistical calculations of the various sole source samples will vary depending upon the specific allelles present at each locus. Instead of stating a single statistic, the statistical calculations for each sole source sample will be broken down into four categories representing separate populations, i.e. African American, 9 Caucasian, South Eastern Hispanic, and South Western Hispanic. This is easily illustrated by a single piece of evidence taken from a blood stain on a shirt recovered from the crime scene. The blood stain was determined to be a sole source evidence sample that matched defendant’s DNA profile at 13 loci. The statistical calculation for this evidence sample is stated as follows: POPULATION RARITY OF DNA PROFILE AFRICAN AMERICAN 1 / 26,000,000,000,000,000,000 CAUCASIAN 1 / 870, 000,000,000,000,000,000 SOUTH EASTERN HISPANIC 1 / 1,000,000,000,000,000,000,000 SOUTH WESTERN HISPANIC 1 / 4,000,000,000,000,000,000,000 When the probability calculation results in an estimate that is more than one in 260 billion, it will be reported in the above format. However, when the probability calculation results in an estimate that is more rare than one in 260 billion, the government’s expert will opine that either the defendant or the decedent (where appropriate) was the source of the detected DNA, to a reasonable degree of scientific certainty. The government’s DNA expert will also discuss mixture samples. In some of the mixture samples, the FBI analyst will testify that the major contributor of sample is defendant and decedent can not be excluded as an additional contributor. Conversely, in other mixtures, the FBI analyst will testify that the major contributor is 10 the decedent and the defendant can not be excluded as an additional contributor. Finally, in one mixture, the FBI analyst will testify that the decedent and defendant are both contributors. In all but one of the mixture calculations, the FBI analyst was able to discern a major contributor profile at 13 loci in the mixture and calculate the statistical significance of the profile as if a single source. In one instance, however, the FBI analyst calculated the match probabilities as though he had no information about possible contributors. In that instance, he used a formula known as the combined probability of inclusion.8/ ARGUMENT I. Defendant’s Supplemental Motion In Defendant’s Supplemental Motion, defendant moves under Frye v. United States, 54 U.S. App. D.C. 46, 293 F. 1013 (1923), to exclude the DNA evidence on the grounds that (1) the collection, analysis, interpretation of nuclear DNA under PCR amplification of short tandem repeats is not generally accepted within the relevant scientific community; (2) the Profiler Plus and Cofiler kits used to apply the PCR/STR methods, and the 310 Analyzer used to identify the STRs at the 13 loci, are not generally accepted; (3) the FBI’s method for calculating the probabilities of a random match at 13 loci is not generally accepted; and (4) the FBI’s alleged failure to follow proper procedures, and its failure to conduct blind proficiency tests, make 8/ This formula is discussed at length, infra. 11 the results of the its DNA testing scientifically unreliable (Defendant’s Supplemental Motion at 1-2). The defendant’s supplemental motion should be denied. As the defendant has noted, the arguments he raises include a “significant amount of material” that has been provided to other judges in D.C. Superior Court considering similar challenges (Defendant’s Supplemental Motion at 3-4). In fact, the defendant has submitted to this Court the same arguments and supporting materials that were submitted to the Honorable Robert I. Richter in the consolidated Frye proceedings involving United States v. Orlando Roberts, F-771-01, and United States v. David Veney, F-3986-00, and to the Honorable Ann O’Regan Keary in the Frye proceedings in United States v. Henry Thompson, F-6540-02. In addition, Judge Richter denied the Roberts and Veney motions in a 20-page written order, a copy of which is included in the government’s Appendix II, Tab X. Judge Keary, primarily for the reasons given by Judge Richter, also denied the motion to exclude DNA evidence in the Thompson case. A copy of Judge Keary’s oral ruling of June 17, 2003, is included for the record in Government’s Appendix II, Tab Y. These issues were also raised before the Honorable Judith E. Retchin in United States v. Sidney Smith, F-9156-98, in which the parties adopted the original DNA pleadings filed in this case and requested that the Court issue a ruling based upon those pleadings. Based upon the original DNA pleadings in this case, Judge Retchin denied the defendant’s motion to exclude DNA evidence. A copy of 12 Judge Retchin’s oral ruling of October 23, 2001 is included for the record in Government’s Appendix I, Tab S. Likewise, the government will put forth in this opposition arguments that are the same or consistent with the previous oppositions filed, and provide to this court the same exhibits. The defendant has requested that this court not look to a ruling issued by Judge Richter in the Roberts and Veney cases for guidance in this case (Defendant’s Supplemental Motion at 4). The defendant finds fault in the fact that Judge Richter did not conduct an adversarial hearing, but, as the court will see upon review of this opposition, such a hearing is not necessary.9/ 9/ The defendant further argues that “Judge Richter also engaged in impermissible burden-shifting” concerning the issue of independent defense testing of the evidence (Defendant’s Supplemental Motion at 4). That argument also must fail. After finding that the FBI’s methods for analyzing DNA and formulas for calculating random match probability were generally accepted, Judge Richter opined that “The extent to which a defendant could theoretically be prejudiced by the possibility of a false positive result is mitigated by the fact that defendants are now legislatively given the right to independent testing under the IPA.” (United States v. Orlando Roberts, F-771-01, and United States of America v. David Veney, F-3986-00, Order by Judge Robert Richter dated March 25, 2003 at 19 (“Roberts/Veney Order”). Judge Richter so recognized independent testing to be a control on false positives in the context of admissibility (see Id.). Judge Richter is recognizing a safeguard put into place by the D.C. City Council that responds to an argument that the defense made concerning a fear of false positives. Because it concerns the admissibility of evidence, this order does not shift the burden to the defendants. 13 The government notes that there are no new arguments raised in Defendant’s Supplemental Motion. Judge Richter’s, Judge keary’s, and Judge Retchin’s opinions can be utilized by this Court as it conducts its own thoughtful inquiry into the issues raised by the defendant. Moreover, this Court could choose to adopt the record and findings created in those cases, and the government would encourage it to do so. Specifically, this Court should decline to consider the defendant’s complaints about the reliability of the specific kits and instruments used to conduct the DNA analysis, the alleged failure to conduct the tests properly, and the lack of blind proficiency tests because, under Frye, those issues go to the weight of the evidence, not its admissibility. The challenges to the PCR/STR process, and to the process of calculating random match probabilities should be rejected because those processes are scientifically accepted, as courts throughout the country have repeatedly found. The District of Columbia Court of Appeals (“DCCA”) has not addressed the admissibility of DNA profile evidence since United States v. Porter, 618 A.2d. 629 (D.C. 1992) (“Porter II”).10/ When Porter II was decided, the use of DNA typing for forensic purposes was still in its nascent stage. Ours was among the first jurisdictions 10/ In Porter II, the DCCA reviewed, and remanded, a decision excluding the DNA evidence. See United States v. Porter, 1991 WL 319015 (D.C. Super. Ct. Sept. 20, 1991) (“Porter I”). On remand, the DNA evidence was admitted. See United States v. Porter, 1994 WL 742297 (D.C. Super. Ct. Nov. 17, 1994) (“Porter III”). 14 to admit DNA evidence, which at the time was obtained by the restriction fragment length polymorphism (“RFLP”) method. The field of forensic DNA analysis has evolved rapidly. In the 12 years since the Porter opinions, identification by DNA typing has become commonplace; the methods used to perform DNA analyses have become more powerful and more discriminating; the tools used to apply those methods have been standardized; and forensic laboratories have adopted quality assurance standards that reflect the maturation of the field of forensic DNA analysis. The methods applied in this case – PCR amplification, STR typing, analysis by laser fluorescent capillary electrophoresis, and calculation of the match probabilities by the unmodified product rule – reflect that maturation. Because there is no published appellate opinion in this jurisdiction affirming the admissibility of the PCR/STR method and the methods for calculating match probabilities used in these cases, those methods are theoretically proper subjects for a Frye hearing. Nonetheless, this Court does not need to conduct an extensive Frye inquiry. The PCR/STR methods used in this case have been scrutinized by the scientific community and other courts for years, and there is virtually universal agreement that they are generally accepted by the relevant scientific community as reliable for forensic DNA analysis purposes. As a result, DNA evidence derived from those methods is routinely admitted in state and federal criminal cases across the country. Similarly, the FBI's methods for calculating the random match probabilities are generally accepted by scientists, and those statistics are routinely 15 admitted to explain the significance of a match between evidence and suspect samples. Courts that adhere to the Frye standard have also consistently rejected arguments, like the defendant’s, that laboratory proficiency testing procedures and laboratory error rates, or the way in which scientific procedures were performed in an individual case, are issues that affect the scientific acceptance of methods of DNA analysis. II. Defendant’s Cold-Hit Motion In Defendant’s Cold Hit Motion, defendant moves to exclude DNA evidence under Frye on the grounds that the statistical calculation utilized by the government to explain the significance of a DNA profile match that is initially derived from trawling through a DNA database is currently the subject of a “raging debate with scientists divided into three separate camps calling for three starkly different approaches” (Cold Hit Motion at 1). The defendant submits that the Court should exclude the DNA evidence until the alleged controversy is quelled. Id. at 2. Defendant’s Cold Hit Motion is equally meritless. Defendant attempts to concoct a scientific controversy over statistical methodologies when, in reality, the various positions cited by the defendant merely reflect answers to different questions that can arise when calculating the statistical probabilities relating to a particular DNA profile in cases where the defendant is initially found by trawling through a DNA database. As we show, defendant’s argument rests on three false premises: 1) there is only one question that statistics can answer in a cold- 16 hit case; 2) the existence of more than one approach to addressing a question means that none is generally accepted; and 3) a debate about the relative merits of different statistical approaches is the equivalent of the “raging controversy” over the fundamental reliability of those approaches. Importantly, there is no controversy regarding the relevance of the estimated rarity of the DNA profile at issue or the general acceptance and admissibility of the Random Match Probability (“RMP “) to answer that question. The RMP answers the question how rare is a DNA profile in a particular population. The statistical method utilized to calculate RMP is generally accepted and the rarity of a particular DNA profile never changes, regardless of whether a defendant is first identified by searching a database. Thus, the rarity of the DNA profile is always relevant for the trier of fact. In 1992, the National Research Council (NRC I) formulated an approach to statistical calculations in the context of a cold-hit case which understated the rarity of the DNA profile to compensate for technological shortcomings in the methodologies being used to amplify DNA in the early 1990s. Under the NRC I approach, DNA markers used to perform the database search would be discarded and the rarity of the DNA profile would be artificially understated by limiting the RMP calculation to a completely new set of DNA markers. As discussed below, this approach, is outdated and impractical, and deprives the jury of relevant evidence. Hence, this approach should not be followed. 17 The RMP answers only one question that can arise in the context of a coldhit case, i.e. how rare is a DNA profile in a population (Question one). In 1996, the National Research Council (NRC II) issued a second report in which the committee addressed a second question that arises when a suspect is identified through a search of a database, i.e., what are the chances of finding a particular DNA profile in a database of a particular size (Question two). Since this question is different from the one answered by the RMP, it results in a different statistical calculation. NRC II assumes the relevance of Question Two and advocates providing a Database Match Probability (DMP) to answer it. To the extent there has been a scholarly discussion regarding statistical calculations in a cold-hit case, the discussion has centered on the relevance of Question Two. While NRC II assumed the relevance of Question Two, another group of scholars opine that, although the second question might be of interest to academicians, and the NRC II formula is an acceptable way to answer it, that question is the wrong one to address in a criminal case. This group believes, in effect, that no other probability figure needs to be introduced with, or instead of, the RMP. The government intends to introduce only the RMP numbers that represent the estimated frequency with which the DNA profiles found in the evidence samples would be expected to occur in four major population groups. We show, the rarity of the DNA profile is always relevant and presentation of the RMP 18 alone does not risk suggesting potentially prejudicial information about the defendant’s criminal record. Additionally, presentation of the RMP in a cold hit case is consistent with the practice of almost every DNA forensic laboratory in the country. If the Court believes that the likelihood of finding a match to the evidence profile in the offender database is relevant, the government does not object if the defense wants to introduce, in addition to the RMP figures, the DMP figure as calculated by the generally accepted formula recommended by the NRC II committee in 1996. There is no controversy in the relevant scientific community that the use of the NRC II formula is both a correct and conservative (as well as a defendant-favoring) way to answer the second question. III. The law in the District of Columbia. District of Columbia courts evaluate the admissibility of novel scientific evidence by the standard first announced in Frye v. United States, 54 U.S. App. D.C. 46, 47, 293 F. 1013 (1923): where expert testimony is not based on a "wellrecognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs." Id. at 1014. See Bahura v. S.E.W. Investors, 754 A.2d 928, 943 n.15 (D.C. 2000) (affirming that Frye test remains in effect in the District of Columbia).11/ 11/ The defendant incorrectly suggests that this Court would be free 19 to apply the federal standard for admissibility announced in Daubert v. Merrell Dow Pharmaceutical, Inc., 509 U.S. 579 (1993)(Defendant’s Supplemental Motion at 9 n.10). The Court of Appeals has recently affirmed that the Frye test still applies in the District of Columbia, see Bahura, 754 A.2d at 943, and this Court is not free to disregard that binding precedent. 20 “Under Frye, the proponent of a new technology must demonstrate by a preponderance of the evidence that this technology has been generally accepted in the scientific community.” Porter II, 618 A.2d at 633. “‘The issue is consensus versus controversy over a particular technique, not its validity.’” Id. at 634 (quoting Jones v. United States, 548 A.2d 35, 42 (D.C. 1988)). “[T]he focus is primarily on counting scientists’ votes rather than on verifying the soundness of a scientific conclusion.” Jones, 548 A.2d at 42. However, unanimity among scientists is not required. Porter II, 618 A.2d at 634. Contrary to the defendant’s contention (Defendant’s Supplemental Motion at 10-11), the Frye test in this jurisdiction does not condition admissibility on a pre-trial showing that generally accepted techniques have been appropriately applied by the laboratory performing the DNA analysis. “[T]he Frye test ‘begins – and ends – with . . . whether there is general acceptance of a particular scientific methodology, not an acceptance, beyond that, of particular study results based on that methodology.’” Porter II, 618 A.2d at 634 (quoting Porter I, [1991 WL 319015 at *15] (in turn quoting Ibn-Tamas v. United States, 407 A.2d 626, 637-39 (D.C. 1979))). “Any failure by the scientists to adhere to the appropriate procedure is, of course, a proper subject of inquiry, but does not raise an issue which implicates Frye.” Id. at 636.12/ “Once the scientific 12/ Like the District of Columbia, many states apply the Frye general-acceptance standard to the theory and technique of a novel scientific method, but treat the question of whether a generally 21 method is shown to be generally accepted, it is presumptively reliable, although the opponent may challenge the weight of the evidence.” United States v. Bridgett, 120 Daily Wash. Law Rptr. 155, at 1702 (D.C. Super. Ct. Aug. 11, 1992)(citing Williams v. District of Columbia, 558 A.2d 344, 346 (D.C. 1989)). This Court has broad discretion as to the kinds of evidence it may consider in a Frye inquiry. The DCCA has held that “the appellate court, like the trial court, may, and often should, pay attention not only to expert evidence of record but also to judicial opinions in other jurisdictions that have considered the question, accepted technique was properly performed as one of weight, not admissibility. See, e.g., State v. Cauthron, 846 P.2d 502, 507 (Wash. 1993) (conditioning admissibility on absence of error is “inappropriate in jurisdictions utilizing the Frye standard of admissibility [because the] core concern of Frye is only whether the evidence being offered is based on established scientific methodology”). Other states add requirements to Frye’s “general acceptance” standard before admitting novel scientific evidence. This variation is known as “Frye-Plus.” See Taylor v. State, 889 P.2d 319, 325 n.13 (Okla. 1995) (comparing jurisdictions, like Oklahoma, that adhere to a “pure” Frye test with those that apply a “FryePlus” test). See also, e.g., People v. Kelly, 549 P.2d 1240 (Cal. App. 3d 1976) (adopting three-part admissibility standard: Part I incorporates Frye general acceptance test; Part II looks at qualifications of the expert witness; Part III requires proof that the scientific test was performed according to a generally accepted methodology); State v. Nose, 649 N.W.2d 815, 819 (Minn. 2002) (state's Frye-Mack test "asks first whether experts in the field widely share the view that the results of scientific testing are scientifically reliable, and second whether the laboratory conducting the tests in the individual case complied with appropriate standards and controls"); People v. Castro, 545 N.Y.S.2d 985 (N.Y. Sup. Ct. 1989) (applying three-part test that adds requirement that techniques be properly performed). 22 as well as to relevant legal and scientific commentaries in which the technique or test has been scrutinized.” Jones, 548 A.2d at 41 (emphasis added) (upholding trial court's admission of EMIT drug-test results based on judicial notice of evidentiary record compiled in different case, and other court opinions that relied on expert testimony and review of scientific literature). In determining the admissibility of scientific evidence under Frye v. United States, 54 U.S. App. D.C. 46, 293 F.1023 (1923), this Court may take judicial notice of extra-record materials, including expert testimony from other proceedings. See District of Columbia ex rel. J.A.B. v. W.R., Jr., 1991 WL 214204, *7-*8 (D.C. Super. Ct. (1991) (“[E]xpert testimony in other cases, subject to crossexamination, can be probative of the general acceptance of a scientific technique. . . . This Court can also take judicial notice of such expert testimony.”) (citing Jones v. United States, 548 A.2d 35, 36 (D.C. 1988)); see also United States v. Bridgett, 120 Daily Wash. Law Rptr. 155, at 1697 (D.C. Super. Ct. Aug. 11, 1992) (taking judicial notice of expert testimony from proceedings in Porter I). The testimony does not have to come from a court in this jurisdiction to be helpful to the admissibility determination. See Porter I, 1991 WL 319015, *28 n.47 (acknowledging usefulness of expert testimony from admissibility hearing in Ohio federal district court); accord Porter II, 618 A.2d at 631 n.4. Where the scientific literature, and opinions and records from other cases clearly demonstrate the general acceptance of a scientific method, a court 23 may even dispense with live testimony. See, e.g., Porter II, 618 A.2d at 635; Bridgett, 120 Daily Wash. Law Rptr. 155, at 1697.13/ In fact, the Court of Appeals rejected a challenge to the reliability of a scientific procedure used to test sobriety (the Horizontal Gaze Nystagmus Test (“HGN”)) simply by taking judicial notice that “‘the great weight of scientific literature supports [the] reliability [of the HGN] and [that] the majority of jurisdictions around the country have declared HGN testing to be reliable.’” Karamychev v. District of Columbia, 772 A.2d 806, 812 (D.C. 2001) (quoting Schultz v. State, 664 A.2d 60, 74 (Md. App. 1995)). Moreover, although the relevant scientific communities include those of forensic DNA science, molecular biology, and population genetics, “the views of forensic scientists have weight and must be considered.” Porter II, 618 A.2d at 634. IV. DNA analysis by the PCR/STR method is universally accepted in the relevant scientific community and the courts. A. Background of DNA analysis. 13/ Although this Court may take judicial notice of judicial opinions and expert testimony from other hearings on the same issue, we object to, and ask the Court to disregard, the defendant’s citations to opinions and testimony that are not readily available and have not been provided to counsel and the Court for inclusion in the record in this case. See People v. Morganti, 50 Cal. Rptr. 2d 837, 852 n.19 (Ct. App. 1996) (court declines to consider unpublished opinions that have not been made part of record in admissibility proceedings). 24 The theories behind DNA typing are now relatively familiar. See, e.g., Porter II, 618 A.2d at 632; Bridgett, 120 Wash. L. Rptr. 155, at 1697, 1700.14/ The nucleus of most cells in a human body contains 23 pairs of worm-like microscopic bodies called chromosomes, each pair containing one chromosome inherited from the mother and one from the father. At the core of each chromosome is a long thin strand of the molecule deoxyribonucleic acid, or DNA. In general, the DNA within an individual is identical, but, as with fingerprints, no two individuals have the same DNA (with the exception of identical twins). The DNA molecule looks like a rope ladder twisted into a helix. The sides of the ladder are phosphate-sugar strands. The rungs are pairs of “bases” or nucleotides. There are four types of nucleotides: adenine (“A”), thymine (“T”), guanine (“G”), and cytosine (“C”). A always pairs with T, and G always pairs with C, so there are only four possible combinations of base pairs at a single rung (AT, TA, GC, CG). The entire human DNA makeup, or “genome,” contains approximately three billion base pairs, ninety-nine percent of which are identical among humans. Forensic DNA identification is based on the ability to recognize 14/ For additional information on the theory, methods, and history of forensic DNA testing, see generally Butler, J., Forensic DNA Typing: Biology & Technology Behind STR Markers (Academic Press 2001). 25 differences in selected sequences of base pairs at specific sites, or loci, in the remaining one percent – approximately three million base pairs – which may vary from person to person. A variation in a base-pair sequence at a particular site, or locus, is called an “allele.” There could be several or even hundreds of variations, or alleles, at a locus. Typically, the DNA sites chosen for forensic examination contain highly variable, non-functional DNA sequences that are repeated over and over again next to each other. These repeat regions are called “variable number of tandem repeats” (“VNTRs”). One individual can be distinguished from another by the number of tandem repeats found at these VNTR sites. B. The PCR process. As a general matter, polymerase chain reaction (“PCR”) is merely a method of amplifying, or copying, a small DNA sample so that it can be easily analyzed using other techniques. Thus, by itself, PCR “is not an analytic tool[.] [R]ather [PCR] facilitates forensic (as well as other) applications by allowing a scientist to take a sample of DNA [that] would generally be insufficient to detect the characteristics of the DNA[] and amplify it until enough copies are available for further analysis.” Office of Technology Assessment, Congress of the United States, Genetic Witness: Forensic Uses of DNA Tests, at 48 (July 1990)(characterizing PCR as “molecular photocopying”). As explained in a 1996 report by the National Research Council (“NRC”): 26 The [PCR] process is similar to the mechanism by which DNA duplicates itself normally. [It] consists of three steps. First, each double-stranded segment [of DNA] is separated into two strands by heating.15/ Second, these single-stranded segments are hybridized with primers, short DNA segments (20-30 nucleotides in length) that complement and define the target sequence to be amplified.16/ Third, in the presence of the enzyme DNA polymerase, and the four nucleotide building blocks (A, C, G, and T), each primer serves as the starting point for the replication of the target sequence. A copy of the complement of each of the separated strands is made, so that there are two double-stranded DNA segments. This three-step cycle is repeated, usually 20-35 times. The two strands produce four copies; the four, eight copies; and so on until the number of copies of the original DNA is enormous. The main difference between this procedure and the normal cellular process is that the PCR process is limited to the amplification of a small DNA region. Committee on DNA Technology in Forensic Science, National Research Council, The Evaluation of Forensic DNA Evidence (1996) (“NRC II”), Appendix II, Tab A, at 69-70. 15/ The strands are separated lengthwise down the middle between the two complementary nucleotides that form each base pair – a process often described as “unzipping.” 16/ If the primer reads ACTT, then it will attach itself to a segment that reads TGAA. A more detailed description of how the PCR hybridization process works can be found in Butler, J. Forensic DNA Typing, at 39-52; see also, e.g., State v. Lyons, 863 P.2d 1303, 1307-08 (Or. App. 1993), aff’d, 924 P.2d 802 (Or. 1996). 27 The defendant asserts that “[e]xcept in observing the high risk of contamenation from multiple amplification cycles, he "does not take issue with the PCR process as a method of amplifying nuclear DNA for analysis purposes" (Defendant’s Supplemental Motion at 6 n.9).17/ The overwhelming weight of scientific and legal opinion finds PCR to be generally accepted in the relevant scientific community, and the results of PCR-based DNA typing are today routinely admitted in courts throughout the country. As early as 1994, courts deciding the admissibility of evidence derived from PCR-based DNA methods found the PCR process to be “routine in many settings,” and “used by nearly every molecular biologist in the world.” State v. Russell, 882 P.2d 747, 765 (Wash. 1994) (en banc) (describing various non-forensic applications of PCR-based DNA 17/ Other courts have reviewed alleged deficiencies in PCR-based DNA typing systems, such as the risk of contamination, and concluded that “the potential testing problems that [defendant] cites are either detectable or preventable.” State v. Russell, 882 P.2d 747, 765, 767-68 (Wash. 1994) (en banc) (rejecting challenges to PCR admissibility based on the alleged unreliability caused by differential amplification, misincorporation, and contamination); see also Commonwealth v. Sok, 683 N.E.2d 671, 679-81 (Mass. 1997) (identifying 20 potential problems that “might inhere in any PCR analysis,” but concluding, nonetheless, that “PCR-based testing . . . is a scientifically valid means of comparing DNA profiles”), abrogated on other grounds, In re Canavan, 733 N.E.2d 1042 (Mass. 2000); State v. Tankersley, 956 P.2d 486, 492 (Ariz. 1998) (en banc)(the risk of contamination "has not rendered PCR-based techniques unacceptable by the scientific community. In fact, erroneous amplification is far more likely to result in the false exclusion, not inclusion, of a suspect. . . . Moreover, the possibility of contamination ‘may present an open field for crossexamination.'") (internal quotations and citation marks omitted)(emphasis added). 28 typing methods). The acceptance of PCR theory and technology in forensic and non-forensic DNA analysis was well documented in the scientific literature. See, e.g., State v. Lyons, 863 P.2d 1303 1310 (Or. App. 1993) (taking judicial notice of a bibliography listing some 4,000 scientific articles and publications relating to PCR-based DNA analysis). In addition, “PCR-based DNA analysis has been used to clear persons suspected of a crime and to exonerate defendants who have been wrongfully convicted.” Sok, 683 N.E.2d at 674 n.8. The general acceptance of PCR-based techniques was affirmed by the National Academy of Science's National Research Council (“NRC”), which conducted two studies devoted exclusively to the use of DNA technology for forensic purposes. In 1990, the NRC convened a special committee of molecular biologists, population geneticists, and other scientists and experts, which concluded after two years of study that “the theory of PCR analysis . . . is scientifically accepted and has been accepted by a number of courts.” See Committee on DNA Technology in Forensic Science, National Research Council, DNA Technology in Forensic Science, at 70 (1992) (“NRC I”) (Appendix II, Tab B).18/ Four years later, the NRC observed in a follow-up report that much more 18/ The conclusions of the NRC were central to the Court of Appeals' decision in Porter to admit DNA evidence with some form of random probability calculation. Porter II, 518 A.2d at 642 n.24, 643 n.26 (calling NRC report "authoritative" and "definitive"). The Court observed that the NRC’s conclusions regarding the reliability of forensic DNA typing and appropriate methods for calculating random match probabilities could “easily be equated with general acceptance of those methodologies in the relevant 29 had been learned about PCR-based testing, and reaffirmed its acceptance as a reliable technology for DNA profiling. See NRC II, Appendix II, Tab A, at 2, 23, 26. In light of the NRC’s endorsement of the PCR method, and the extent to which PCR has been subject to scientific examination, it is not surprising that “[t]he majority of courts . . . that have considered the issue have held that DNA evidence derived from the PCR testing method satisfies the standards for admissibility under either Frye or [Colorado Rule of Evidence 702's relevance/reliability standard].” People v. Shreck, 22 P.3d 68, 79 n.15 (Colo. 2001)(collecting cases); People v. Butterfield, 27 P.3d 1133, 1143 (Utah 2001) (same). C. The current generation of PCR-based analysis: STRs. 1. Background. Because the PCR method alone merely copies DNA segments, a second technique must be used to identify or “type” certain sites on the DNA strand for use in comparison to a second DNA sample to determine if there is a match. Since 1992, when Porter II affirmed the admissibility of the RFLP method, at least four PCR-based typing methods have come into widespread use. The most scientific community.” Id. at 643 n.26. See also Bridgett, Daily Wash. Law Rptr. 155, at 1702 (bodies such as the “serve[] as a voice of reason not only with respect to divergent views postulated by the various experts who [testify] . , but also with respect to the divergent views expressed in numerous scientific articles submitted by the parties”). 30 120 NRC the . . the current method – which examines DNA segments called short tandem repeats (“STRs”) – was used in the Jenkins case.19/ 19/ The other three techniques, which are predecessors to the PCRSTR methods used here, involve 1) a single locus called DQ-Alpha (also denoted “Dqα,” “DQ-α,” or “DQA1"); 2) five loci (low-density lipoprotein receptor (LDLR), glycophorin A (GYPA), hemoglobin B gamma globin (HBGG), D7S8, and group-specific component (GC)) that are analyzed through the AmpliType® Polymarker PCR amplification and typing kit (“PM”); and 3) another single locus named D1S80, which is a 16-base-pair VNTR. See, e.g., Sok, 683 N.E.2d at 679-81 (describing various systems). Courts have routinely admitted DNA match results from these three testing methods based on findings that the methods are generally accepted in the scientific community. Id. 31 The PCR/STR method represents, to date, the most powerful of the analytic tools used for DNA typing. Just like the RFLP process accepted in Porter, STR technology examines non-functional DNA sites, or loci, where there are basepair sequences that repeat themselves in varying numbers. Unlike the RFLP process, which examines alleles that range from 500 to 10,000 base pairs in size, the STR process focuses on tandem repeats in which the alleles are much smaller, ranging from 100 to 400 base pairs.20/ Once DNA for analysis is obtained, the target STRs in that DNA are identified and marked during the first cycle in the PCR amplification process. A 20/ Much of the theory and technology behind the RFLP method is similar to the theory and technology behind the PCR/STR method, although the STR method is much simpler. RFLP testing involves a seven-step process: removal of the DNA from the evidence sample ("extraction"); use of an enzyme-based process to separate targeted sequences of base pairs from the DNA molecule ("digestion"); use of a gel-electrophoresis process to further separate the targeted sequences by length ("separation by gel electrophoresis"); severing of the targeted sequences lengthwise into complementary halves and transfer to a different medium for further processing ("Southern transfer"); application of radioactively charged single-stranded complementary segments of DNA, called probes, to identify the target DNA fragments ("hybridization"); placement of radioactivelymarked segments between x-ray film to create visual patterns of bands, or autoradiographs, unique to each targeted DNA fragment ("autoradiography"); and comparison, first by a trained analyst, then by a computer process, of the autoradiographs produced by the evidence samples and the known samples (e.g., from a victim and a suspect) to determine whether there is a match ("interpretation"). Porter I, 1991 WL 319015, at *3-*5. RFLP testing cannot identify the exact number of VNTRs, so the length must be estimated, and comparisons must be made using “match windows” which compensate for known measurement error. Id. at *5. If fragments from a suspect’s DNA sample fall within the same match window as fragments of DNA from the evidence sample, then the samples will be called a match. 32 piece of DNA is heated until it splits into two strands, then short pieces of synthetic DNA, called primer pairs, are added to the DNA. Each primer pair will attach itself at either end, and outside of, an area of DNA known to contain a specific STR. Chemicals are then added to induce replication, or copying of only the desired STR sites. This process is repeated many times, until an enormous number of copies of the targeted STRs are obtained from the original source of DNA. The simplest way to mark and copy an STR is to use a “monoplex” process, in which only one primer pair is used during the PCR cycle and, thus, only one targeted STR locus is amplified. More useful, however, is a “multiplex” process, during which several primer pairs are used simultaneously and multiple targeted STR loci are amplified concurrently. In a multiplex process, the various primers for the targeted loci contain different colored fluorescent dyes, so that alleles at different STR loci can be distinguished from one another later in the process. The two commercial STR kits used in this case together identify 13 separate STR loci on the DNA molecule. Profiler Plus enables the amplification and typing of the following loci: D3S1358, D7S820, vWA, FGA, D8S1179, D21S11, D18S51, D5S818, and D13S317. Cofiler enables amplification and typing of the loci D16S539, THO1, TPOX, and CSF1PO, as well as two loci also targeted by Profiler Plus, D3S1358 and D7S820.21/ By testing the same two loci, the two kits act as 21/ The labels usually identify the chromosome and the specific 33 internal accuracy controls for every analysis in which they are both used. Those combined loci correspond to the 13 loci used in a national databank containing DNA profiles of persons convicted of crimes, the Combined DNA Index System or “CODIS,” which Congress authorized in 1994. See 42 U.S.C. § 14132 (West 2000). The 13 loci were selected for CODIS through a collaborative process involving FBI and 21 forensic laboratories based on underlying genetic research that had validated these sites for the variability and genetic independence necessary to allow them to be used for identification purposes. See United States v. Trala, 162 F. Supp. 2d 336, 343 (D. Del. 2001). The Profiler Plus and Cofiler kits were developed to identify the CODIS loci. After the target STR loci are amplified through PCR, the different STR segments of DNA are separated by a process called capillary electrophoresis (“CE”). Like the gel-electrophoresis process approved in Porter I, see 1991 WL 319015, *4, the CE process uses an electric current to draw the amplified DNA through a sieve-like medium. A series of DNA fragments of known size, called “standards,” are marked with a red dye and drawn through the medium along with the unknown DNA fragments. The length of each STR can be used to determine how many repeats are present in a fragment and thus to determine an allele designation. The length of the STR can be calculated by how long it site on the chromosome at which these particular STRs can be found. 34 takes it to move through the medium in comparison to the time it takes the size standard to move through the gel. Shorter STRs move more quickly. Instead of running the DNA fragments across a flat bed of gel, as in RFLP, the CE process in the instant cases was conducted with a machine called the 310 Analyzer, which draws the DNA fragments through a narrow capillary tube filled with a polymer. Also, instead of recording the position of the fragments using radioactive probes to create autoradiographs, as in RFLP, the 310 Analyzer sends a laser beam through a small window in the capillary, which illuminates the fluorescent dyes attached to the target sequences and the red size standard as they pass the window. The Analyzer distinguishes the various dye colors by the differences in fluorescent wavelengths, and records how long it takes each unknown fragment and the known size standard to travel through the gel. After the 310 Analyzer completes the fragment separation and collection of data on the colors and relative speeds of the DNA fragments, that information is analyzed by two computer programs, and the results are reported on a graph, or electropherogram. Typically, an electropherogram will show three or four rows of lines with a series of clusters of peaks of varying heights and colors on them.22/ Each cluster represents an allele site, or locus. If sufficiently high, each 22/ Sample electropherograms depicting the results of the Profiler Plus and Cofiler tests can be found at pp. 9-3 to 9-6 of the FBI Laboratory’s “Short Tandem Repeat Analysis Protocol” (hereafter 35 peak within a cluster, or locus, will represent an allele. Each peak is labeled with a number representing the number of repeats found at that allele. The analyst interprets the electropherogram to determine, among other things, whether the peaks shown represent true alleles, and whether they indicate one or more contributors of DNA. In contrast to RFLPs, whose length could only be approximated, STRs can be measured down to the number of repeated base pairs constituting each allele. For each individual, the genotype at a locus will be described by two numbers, one each for the number of repeats on each of the paired chromosomes (maternal and paternal) within the allele (e.g., “16, 18"). Once the allele pairs are determined for each of the 13 separate loci identified by the Profiler Plus/Cofiler/310 Analyzer system, the evidence sample profiles are compared to the known sample profiles to see if there is, or is not, a match. As is true with all of the DNA typing methods, if a profile match is found, then the analyst goes on to calculate the random-match probabilities. See infra at Section VII. “FBI Protocols”) (Appendix II, Tab F). 36 2. The STR method is generally accepted. Because STR-based typing is of more recent vintage than the PCR method, the courts began to address the admissibility of results from STR testing more recently. Nonetheless, since 1997, when the first appellate decision to address STR typing appeared, the overwhelming majority of courts to address this issue has found the STR process to be well accepted in the scientific community.23/ Those courts rely in large part on the NRC’s endorsement of the STR method in its 1996 report (“NRC II”). Judge Richter recognized this acceptance in the Roberts and Veney cases concluding that “The government has met its burden in establishing that STR-based typing is generally accepted.” (Roberts/Veney Order at 13). Although the authors of NRC II acknowledged that the use of STRs was relatively new in 1996, they expressed no reservations about the scientific acceptance of the underlying principles and techniques. In fact, the report emphasized the scientific acceptance of all of the profiling techniques considered, which included PCR-based STR analysis: 23/ Following an extensive admissibility hearing at which seven expert witnesses testified, one Michigan trial judge concluded: “The use of STRs is so accepted that journals will no longer accept articles regarding their use; there is nothing left to explore in the judgment of the world’s scientists.” People v. Phillips, Case No. 00-02025-FC, Op. at 7 (Mich. Cir. Ct., Kent Co., Oct. 20, 2000) (Appendix I, Tab L). 37 We affirm the statement of the 1992 report that the molecular technology is thoroughly sound and that the results are highly reproducible when appropriate quality-control methods are followed. The uncertainties that we address in this report relate to the effects of possible technical and human errors and the statistical interpretation of population frequencies, not to defects in the methodology itself. NRC II at 23. Similarly, the National Institute of Standards and Technology “has determined that there are several advantages of using STRs over conventional techniques, and that the use of STRs for genetic mapping and identity testing has become widespread among DNA typing laboratories.” Shreck, 22 P.3d at 80 (citing John M. Butler & Dennis J. Reeder, “Short Tandem Repeat DNA Internet Database,” http://www.cstl.nist.gov/biotech/strbase/intro.htm). Appellate courts in at least ten states have found DNA profile evidence from PCR/STR multiplex testing to be admissible, and many have admitted the results of testing conducted by the same kits used in this case - the Perkin Elmer Profiler Plus and/or Cofiler kit, and the 310 Analyzer. See State v. Traylor, 656 N.W.2d 885 (Minn. 2003)(reversing decision to exclude DNA evidence; state supreme court finds that PCR/STR testing using the Profile/Cofiler kits and the 310 Analyzer is generally accepted); Troxell v. State, 778 N.E.2d 811, 815-816 (Ind. 2002)(affirming trial court’s determination that PCR/STR testing by FBI laboratory was “reliable and generally accepted in the relevant scientific community”); People v. Hill, 107 Cal. Rptr. 2d 110, 117-18 (Ct. App. 2001) (affirming trial court's 38 refusal to conduct hearing on admissibility of results of Profiler Plus test kit, because kits are just a means of implementing PCR/STR technique which has already achieved general acceptance (citing People v. Allen, 85 Cal. Rptr. 2d 655, 659-60 (Ct. App. 1999) (STR testing found generally accepted under California's Frye-Plus standard));24/ Shreck, 22 P.3d at 81 (adopting Colorado Evidence Rule 702 relevance/reliability test in place of Frye; finding STR method reliable; no need to examine particular Profiler kit); Yisrael v. State, 827 So. 2d 1113, 1114-1115, (Fla. Ct. App. 2002) (affirming admissibility under Frye of results of Profiler Plus and 310 Genetic Analyzer); State v. Lemour, 802 So. 2d 402, 40508 (Fla. Ct. App. 2001) (admitting, under Frye standard, results of PCR/STR triplex kit created by private lab), review denied, 821 So. 2d 297 (Fla. 2002); Commonwealth v. Rosier, 685 N.E.2d 739, 743 (Mass. 1997) (PCR-based tests, including STR, are scientifically valid); Watts v. State, 733 So. 2d 214, 222-23, 227 (Miss. 1999) (affirming, under a Frye-Plus test, trial court’s admission of PCR-based typing, which included analysis by an STR triplex system); State v. Salmon, 89 S.W.3d 540, 542-545 (Mo. Ct. App. Oct. 8, 2002) (upholding, under Frye, results of PCR/STR-based tests, based on testimony of one expert and judicial notice of 24/ The PCR/STR test kit at issue in Hill was the Profiler Plus kit used in the instant case. The court found no legal significance in the fact that the Allen opinion, on which the Hill court relied for the general acceptance of STR testing, decided the admissibility of a different kit that tested fewer loci (a kit by Promega that tested only four loci). Hill, 107 Cal. Rptr. 2d at 117-18. 39 appellate decisions from other states admitting results from similar PCR/STR tests); State v. Jackson, 582 N.W.2d 317, 325 (Neb. 1998) (upholding admission of results of PCR/STR testing under Frye-Plus standard); Butterfield, 27 P.3d at 1141-45 (reliability test under state rules of evidence; taking judicial notice of general acceptance of PCR/STR analysis, including acceptance of the Profiler Plus and 310 Genetic Analyzer, as expressed in scientific publications and opinions from other jurisdictions). In addition to those recent appellate decisions, the trial courts that have examined PCR/STR-based DNA analyses have overwhelmingly found those methods to be generally accepted in the scientific community. As indicated below, most of those courts applied the Frye general-acceptance standard, or a Frye-Plus standard, and many admitted the results of the Perkin Elmer Profiler Plus and/or Cofiler kits, and the 310 Analyzer. Copies of unpublished opinions listed below are provided in Appendix I, at the tab indicated.25/ State v. Lynch, CR 98-11390, Ruling at 6 (Ariz. Super. Ct., Maricopa Co., Aug. 20, 1999)(Frye jurisdiction [see Tankersley, 956 P.2d at 491]; “DNA analysis using STR testing [Profiler Plus and Cofiler] is generally (actually universally) accepted in the forensic science community”) (Tab A); People v. Hunt, Case No. SA034500, Transcript of Oral Ruling (Cal. Super. Ct., Los Angeles Co., Oct. 24, 2000) (Frye-Plus; admitting evidence from Profiler Plus and Cofiler)(Tab B); 25/ The unpublished opinions appear in Appendix I as they are listed in the text: in alphabetical order by state, and, within each state, in reverse chronological order. 40 People v. Elizarraras, Case No. 50651, Ruling (Cal. Super. Ct., Tulare Co., Oct. 13, 2000)(Frye-Plus; general acceptance hearing not required for kits but ruling in the alternative that results from Profiler Plus and 310 Analyzer admissible) (Tab C); People v. Moevao, Case No. 168277, Opinion and Order (Cal. Super. Ct., San Francisco Co., July 25, 2000)(Frye-Plus; admitting results derived from Perkin Elmer Blue and Green I kits, which were later incorporated in Profiler Plus and Cofiler kits, laser fluorescent capillary electrophoresis, and 310 Analyzer; proceedings involved 25-day hearing with 8 witnesses and 60 exhibits, and resulted in over 2300 pages of transcripts) (Tab D); People v. Hill, Case No. 232982, Ruling (Cal. Super. Ct., Santa Barbara Co., April 18, 2000)(Frye-Plus; finding admissible PCR-based STR tests using unnamed kits developed by Perkin Elmer) (Tab E); aff'd, 107 Cal. Rptr. 2d 110 (Ct. App. 2001); People v. Bertsch, et al., Case No. 94FO7295, Transcript of Oral Ruling (Cal. Super. Ct., Sacramento Co., Oct. 20, 1999)(Frye-Plus; holding that California law does not require kits or instruments used to implement PCR/STR technique to meet general-acceptance standard, but, if it did, Profiler Plus and 310 Analyzer meet that standard) (Tab F); State v. Flores, No. 99CR2960, Transcript of Oral Ruling (Colo. Dist. Ct., City and Co. of Denver, Dec. 28, 2000), (Frye test; admitting results of Profiler Plus and 310 Analyzer) (Tab G); State v. Flores, No. 99CR2022, Transcript of Oral Ruling (Colo. Dist. Ct., Adams Co., Oct. 20, 2000)(Frye test; admitting results of Profiler Plus) (Tab H);26/ State v. Grant, 32 Conn. L. Rptr. 30, 2002 WL 853627 (Conn. Super. Ct. April 9, 2002) (Daubert standard; court finds Profiler Plus/Cofiler results admissible based on judicial notice of decisions from other 26/ There are two decisions, by two different Colorado trial courts, involving the defendant Reynaldo Luna Flores. 41 jurisdictions and, in the alternative, after application of Daubert factors to evidence); State v. Yisrael, No. 99-20176CF10A, Ruling at 3-4 (Fla. Cir. Ct., Broward Co., Aug. 8, 2000)(“DNA typing using the Profiler Plus Kit with the 310 Genetic Analyzer has now gained general acceptance by the relevant scientific community and thus meets the Frye standard”) (Tab I), aff'd, 2002 WL 31355273 (Fla. Ct. App. 2002); Commonwealth v. Gaynor, Case No. 98-0965-0966, Memorandum of Decision (Mass. Super. Ct., Hampden Co., Apr. 13, 2000)(may use Frye general acceptance or modified Daubert test;27/ admitting results of Profiler Plus, Cofiler, and 310 Analyzer) (Tab J); People v. Kopp, et al., Case No. 00-04014-FH, et al., Opinion at 11 (Mich. Cir. Ct., Kent Co., Oct. 20, 2000)(Frye-Plus jurisdiction; expert witnesses “establish by what can only be called overwhelming evidence that the PCR STR methodology as well as the Perkin-Elmer CoFiler and ProFiler kits utilizing the 310 Genetic Analyzer . . . have received broad and, in fact, universal acceptance in the scientific community”) (Tab K), aff'd 2002 WL 31058350 (Mich. Ct. App. 2002) (assuming, arguendo, there was error, error was harmless given weight of evidence); People v. Phillips, Case No. 00-02025-FC, Opinion (Mich. Cir. Ct., Kent Co., Oct. 20, 2000) (Frye-Plus; finding that prosecution and defense witnesses agree that PCR-based STR analysis is generally accepted, and rejecting defense expert’s opinion that results from 310 Genetic Analyzer are not) (Tab L); People v. Cavin, Case No. 00-4395-FY, Opinion and Order (Mich. Cir. Ct., Lake Co., Oct. 18, 2000) (Frye-Plus; after considering 27/ Massachusetts courts are allowed to evaluate the admissibility of novel scientific evidence under either the Frye generalacceptance standard or a modified relevance/reliability standard as expressed in Daubert, 509 U.S. 579. See Commonwealth v. Lanigan, 641 N.E.2d 1342 (Mass. 1994) (noting that in most cases general acceptance will be the significant and “often the only issue”). 42 testimony of seven expert witnesses, court finds results of Profiler Plus and 310 Analyzer admissible)(Tab M);28/ State v. Roman Nose, 649 N.W.2d 815 (Minn. 2002)(trial court originally admitted the DNA evidence without conducting an evidentiary hearing. The state supreme court later held that, under Minnesota law, such a hearing was required. On remand, after receiving numerous articles and other exhibits, and hearing the testimony of thirteen witnesses, some of which addressed topics beyond the issues relevant to general acceptance, the trial court in Roman Nose concluded that “it is generally accepted in the relevant scientific community [including scientists experienced in molecular genetics, population genetics, and forensic science] that” 1) PCR/STR amplification and typing methods, 2) PCR/STRbased capillary electrophoresis analysis of DNA, and 3) application of the PCR/STR method of analysis by the Profiler Plus and Cofiler kits, the 310 Genetic Analyzer, and the Genescan and Genotyper software “produce results that are scientifically reliable as accurate, and that the combined methods of PCR/STR analysis and statistical formulas used in the field of population genetics accurately determine certain frequencies necessary to understand the significance of the results of DNA profile matches or non-matches” (see Order dated January 3, 2003, Appendix VI Tab B at 2729). The trial court also concluded that using a random match probability calculation for a single-source DNA sample, or, in some circumstances, for a mixed DNA sample (i.e., a sample containing DNA from more than one contributor) where the profile of an individual contributor can be discerned, is generally accepted. (Appendix VI at Tab B) 28/ The Kopp, Phillips, and Cavin cases from Michigan were decided by different judges based on the same evidentiary hearing at which six experts testified that the PCR and STR methods, and the Perkin Elmer kits and 310 Analyzer, all were generally accepted by the scientific community. See Kopp, Op. at 2 (Appendix I, Tab K). Under Michigan law, the judges who presided jointly over this consolidated record were required to find that the experts were disinterested witnesses before they could give their testimony any weight. See Kopp, Op. at 3-4, 11 (Appendix I, Tab K); Phillips, Op. at 2-3, 7-8 (Appendix I, Tab L); Cavin, Op. at 17-18 (Appendix I, Tab M). The Cavin opinion, at 18-43, summarizes the testimony. 43 State v. Kirkendahl, Case No. 00044987, Order (Minn. Dist. Ct., 4th Dist., Hennepin Co., Jan. 16, 2001) (Frye-Plus jurisdiction; although theory and techniques of PCR/STR analysis are all that must pass general acceptance test, “tools” applying those techniques – Profiler, Cofiler, and 310 Analyzer – also pass State’s additional reliability standard) (Tab N);29/ State v. Dishmon, File No. 9947345, Order and Memorandum at 8-9 (Minn. Dist. Ct., 4th Dist., Hennepin Co., Mar. 3, 2000) (Frye-Plus; Frye does not require that kits and instruments be generally accepted in scientific community, but if it did, Profiler Plus, Cofiler, and 310 Analyzer would meet that standard; findings based on testimony from 11 experts) (Tab O); State v. Keightley, Case No. 30R090100013, Docket Entry (Mo. Cir. Ct., Webster C., May 22, 2002) (Frye jurisdiction; finding hearing not necessary, and admitting results of Profiler Plus/Cofiler test kits based on review of cases and scientific articles) (Tab P); State v. Staples, Case No. CR1999-03841, Order (Mo. Cir. Ct., Jackson Co., Nov. 15, 2000)(Frye jurisdiction; admitting PCR-based STR analysis conducted by FBI using an unnamed Perkin Elmer system) (Tab Q); State v. Whittey, Case No. 00-S-273-275, 942, Order (N.H. Super. Ct., Merrimack Co., May 22, 2001) (under Frye-Plus test, admitting results from Profiler Plus/Cofiler/310 Genetic Analyzer kits and system) (Tab R); State v. DeLoatch, 804 A.2d 604 (N.J. Super. Ct. 2002) (upholding, under Frye, admission of results of Profiler Plus/Cofiler/310 Genetic Analyzer based on testimony of one expert and judicial notice of opinions from other jurisdictions); 29/ The Kirkendahl court did not hear live testimony, but relied instead on the transcripts of the testimony from 11 expert witnesses in the Dishmon case, infra. The Minnesota Supreme Court has recently disapproved of the practice of relying exclusively on judicial notice of extra-record material in determining the general acceptance of novel scientific techniques under that state's twopart Frye-Plus admissibility standard. See State v. Nose, 649 N.W.2d 815 (Minn. 2002). 44 People v. Owens, 725 N.Y.S.2d 178 (Super. Ct., Monroe Co., 2001) (denying, without a Frye hearing, motion to exclude DNA evidence because admissibility of PCR/STR test results well established; here tests performed by Profiler Plus and Cofiler kits). Scientific and legal opinion weighs so heavily in favor of the admissibility of the PCR/STR method of DNA typing that this court may, with confidence, hold that the PCR/STR method is admissible under Frye by taking judicial notice of those authorities. See Karamychev, 772 A.2d at 812; Porter II, 618 A.2d at 635. V. The defendant’s arguments against the DNA typing methods in this case go to weight of the evidence, not its admissibility. Despite the weight of authorities that find the PCR and STR methods admissible under Frye, the defendant contends that this Court cannot admit evidence derived from those methods unless it also finds that the kits and instruments by which those methods are applied, and the actual application of those techniques by the individual FBI analyst in the Jenkins case, is also generally accepted in the scientific community. The defendant asserts that the DNA evidence in this case is inadmissible because Perkin Elmer does not disclose the primer sequences used in its Profiler Plus and Cofiler kits; the FBI lab does not conduct blind proficiency tests of its examiners and does not "track and disclose (or estimate) error rates"; and the methodology used by the individual analyst in his analyses of the DNA samples was unreliable. The defendant is wrong as a matter of law and fact. 45 A. The Court does not have to find that the tools used to apply the PCR/STR technique are generally accepted; even if it did, the kits and instrument used in these cases are generally accepted. The defendant’s argument – that the DNA test results cannot be found generally accepted under Frye because the Perkin Elmer company has not publicized the primer sequences used in Profiler Plus and Cofiler – has been rejected by courts across the country. 1. No inquiry is required under Frye. This Court should decline to consider the defendant’s argument primarily because, under Frye, the reliability of the tools used to apply generally accepted scientific methods goes to the weight of the evidence, not its admissibility. For example, appellate courts in California and Colorado reversed trial courts that had excluded DNA evidence on the grounds that the kits had not been generally accepted. See People v. Bokin, Case No. 168461, Order (Super. Ct., San Francisco, May 6, 1999) (Frye-Plus standard); People v. Shreck, Case No. 98CR2475, Ruling and Order (Dist. Ct., Boulder Co., Apr. 12, 2000) (Frye standard), overruled, People v. Shreck, 22 P.3d 68 (Colo. 2001). The appellate courts reversed the Bokin and Shreck trial courts in large part because the trial courts extended the Frye test beyond its scope when they began to scrutinize the tools used to apply otherwise accepted scientific techniques. The Colorado Supreme Court overturned the trial court’s ruling in Shreck, saying, in part, that 46 the trial court erred by subjecting the Profiler Plus and Cofiler kits to an admissibility analysis because “questions as to the reliability of the particular type of multiplex kit go to the weight of the evidence, rather than its admissibility.” Shreck, 22 P.3d at 81 (citing State v. Russell, 882 P.2d 747, 768 (Wash. 1994) (en banc)).30/ Similarly, a California appellate court has effectively rejected the trial court opinion in Bokin by affirming another trial court’s decision to admit the results of the Profiler Plus kit without conducting any hearing on admissibility. See People v. Hill, 107 Cal. Rptr. 2d 110 (Ct. App. 2001) (affirming Hill, No. 232982, Appendix I, Tab E). The court of appeals held that California law did not require the trial court to find that the Perkin Elmer kit used to perform the DNA analysis was generally accepted because The Profiler Plus kit does not embrace new scientific techniques. To the contrary, it uses the PCR and STR testing methods which are generally accepted by the scientific community. It is just more sophisticated because it examines a greater number of genetic markers. Id. at 118. 30/ In reversing the trial court’s decision in Shreck, the Colorado Supreme Court took the occasion to change from a Frye admissibility standard to one based on Colorado Rule of Evidence 702 which, like its federal counterpart, looks at relevance and reliability. Shreck, 22 P.3d at 78. However, as illustrated by the Russell opinion, the Colorado Supreme Court’s reasoning concerning the scope of a general-acceptance inquiry applies with equal force to a Frye inquiry. 47 Other courts agree that, where methods such as PCR and STR are found to be generally accepted in the scientific community, there is no need to establish the general acceptance of the tools used to implement those methods. See, e.g., Elizarraras, Ruling at 3-4 (Appendix I, Tab C) (where same basic scientific principles are at work, and principles have been accepted by scientists and the courts, kits not subject to general-acceptance test); Lemour, 802 So.2d at 407 (Frye test does not require court to find admissible the noncommercial PCR/STR triplex kit developed by private lab); Kirkendahl, Op. at 15 (Appendix I, Tab N) (“Frye evaluated a lie detector technique, not a specific blood pressure cuff. The principles underlying fingerprint comparison are generally accepted, so a crime lab should not have to show general acceptance of its choice of every new fingerprint powder.”); Russell, 882 P.2d at 768 (Frye test does not require the general acceptance of the means by which a DNA testing method is implemented because “[t]he . . . kit is simply one tool for carrying out generally accepted [PCR-based DNA typing] methodology, and any concerns about its implementation in a given case are matters to be addressed to the trial court.”); cf. United States v. Trala, 162 F.Supp. 2d 336, 346 (D. Del. 2001) (applying Daubert standards, district court finds that "kits do not represent separate part of the typing process, but rather, simply contain materials for beginning of the [already accepted] PCR process").31/ 31/ But cf. Bertsch, Tr. at 20363-65, 20373-74 (Appendix I, Tab F) 48 (kits and 310 Analyzer not subject to general-acceptance test; however, court suggests that “information regarding an individual kit may be relevant to initial general acceptance evaluation of a test such as STRs,” id. at 20365). 49 This Court should decline to review the general acceptance of the Perkin Elmer kits for the same reasons. Frye speaks in broad terms of novel scientific theories and methods. In the Porter cases, neither the trial court nor the Court of Appeals suggested that it was necessary to scrutinize the tools employed to implement the RFLP method of analysis. Each court readily affirmed the rule in this jurisdiction that the Frye inquiry “begins – and ends – with . . . whether there is general acceptance of a particular scientific methodology,” not an acceptance of anything beyond that. Porter I, 1991 WL 319015, *22; accord Porter II, 618 A.2d at 633. It would be inconsistent with this reasoning to subject the kits used in this case to a Frye inquiry. Rather, as other courts have done, the Court should permit the defendant to scrutinize the application of otherwise admissible scientific methods through the ordinary methods of cross-examination and defense evidence at trial. 2. The kits are generally accepted. Even if the general acceptance of the kits were a proper subject for a Frye inquiry, this Court could find the kits admissible based solely on the findings and opinions of the many courts that have held, within the past five years, extensive evidentiary hearings and concluded that the results derived from the Profiler Plus and Cofiler kits and 310 Analyzer are admissible. See, e.g., Traylor, 656 N.W.2d at 893, at 900 (court finds results from Profiler/Cofiler310 Analyzer system both “generally accepted” and “reliable” under state’s Frye-plus 50 admissibility standard); Flores, 12/28/99 Tr. at 161-65 (Appendix I, Tab G)(based on expert testimony and scientific literature, court finds DNA typing using Profiler Plus and 310 Analyzer to be generally accepted); Yisrael, Ruling at 3-4 (Appendix I, Tab I)(“DNA typing using the Profiler Plus Kit with the 310 Genetic Analyzer has now gained general acceptance by the relevant scientific community and thus meets the Frye standard”); Kopp, Op. at 11 (Appendix I, Tab K)(testimony of six expert witnesses “establish by what can only be called overwhelming evidence that the PCR STR methodology as well as the PerkinElmer CoFiler and ProFiler kits utilizing the 310 Genetic Analyzer . . . have received broad and, in fact, universal acceptance in the scientific community”); Owens, 725 N.Y.S.2d at 841 (admitting DNA evidence without a Frye hearing, based in part on finding that "[c]ourts throughout the country have found that the Short Tandem Repeat (STR) DNA profiling, using the AmpF/STR Profiler Plus and Cofiler PCR kits are reliable and generally accepted by the scientific community") (collecting cases); Butterfield, 27 P.3d at 1144 ("the forensic community and those quarters of the scientific community that have used the [Profiler Plus and 310 Analyzer] uniformly assert the system's reliability")(collecting scientific studies).32/ 32/ The defendant’s cursory challenge to the 310 Analyzer and its related Genescan and Genotyper software is wholly baseless. The National Institute of Standards and Technology (“NIST”) considers the method of laser fluorescent capillary electrophoresis ("CE") to be an “established technology.” See NIST, “STR Typing Technology 51 Judge Richter, recognizing this case law, further found the tools used by the FBI to implement the PCR/STR technology were generally accepted, and that the FBI laboratory’s protocols for implementing this technology were minimally reliable and generally accepted. (Roberts/Veney Order at 13). a. Adherence to the TWGDAM guidelines is not required for general acceptance. Review” (Appendix II, Tab H). So do courts throughout the nation. See, e.g., Moevao, Op. at 30-38 (Appendix I, Tab D) (analysis by CE method, particularly as applied by 310 Genetic Analyzer, generally accepted, as shown by expert testimony and 21 scientific articles); Yisrael, Ruling at 4 (Appendix I, Tab I)(finding 310 Analyzer has been subject to peer review because over 30 articles have been published dealing with it or comparable instrument using CE). 52 There is no merit to the defendant’s assertion that the kits cannot be found to be generally accepted because Perkin Elmer did not comply with the validation guidelines promulgated by the Technical Working Group on DNA Analysis Methods (“TWGDAM”).33/ (Defendant’s Supplemental Motion at 21-22). Our courts have never suggested that the scientific acceptance of a particular DNA typing method depends on adherence to the TWGDAM guidelines, and the vast majority of courts to have addressed this question concludes that the TWGDAM guidelines were not intended to be used to restrict the admissibility of DNA evidence. See, e.g., State v. Tankersley, 956 P.2d 486, 493 (Ariz. 1998) (TWGDAM guidelines are not mandatory); Hill, 107 Cal. Rptr. 2d at 118 (rejecting as "without merit" the argument that "TWGDAM guidelines on developmental validation of DNA test kits are mandatory and the sine qua non for admissibility"); Elizarraras, Op. at 6 (Appendix I, Tab C)(“[TWGDAM] guidelines are [not] mandatory; rather, as stated in the Introduction to the 1991 [TWGDAM] Guidelines, failure to comply with each and every guideline, or use of 33/ TWGDAM was created by scientists at the FBI research laboratory in the late 1980s to address the need for quality assurance and quality-control standards for forensic laboratories, which were increasingly using DNA analyses. Porter III, 1994 WL 742297, at *7. TWGDAM recommended procedures for forensic DNA laboratories to follow in such areas as validating new DNA technologies, monitoring analysts’ proficiency, and documenting test results. The first guidelines took effect in 1991. TWGDAM updated them in 1995, primarily to address PCR-based systems. The TWGDAM guidelines were later superseded by guidelines developed by the DNA Advisory Board ("DAB") created by Congress in 1994. See infra at note 33. 53 alternative or equivalent methods does not necessitate a finding of unreliable results”);34/ Moevao, Op. at 26 (Appendix I, Tab D) (“failure to adhere literally to the TWGDAM guidelines does not warrant exclusion of otherwise relevant evidence”); Bertsch, Tr. at 20369 (Appendix I, Tab F)(TWGDAM guidelines “are not conditions precedent that have to be met in order for a court to find general acceptance”); Lemour, 802 So. 2d at 407 (TWGDAM guidelines are merely advisory and, in any event, have been replaced by DAB guidelines); Kopp, Op. at 13 (Appendix I, Tab K)(Frye-Plus law does not require that validation studies meet standards set by a federal committee); Owens, 725 N.Y.S.2d at 183 (adherence to proper protocols and quality-assurance standards goes to weight not admissibility); see also Bridgett, 120 Daily Wash. Law Rptr. 155, at 1703 (noting that NRC called TWGDAM guidelines an “instructive example” of quality-assurance and quality-control guidelines); cf. State v. Jobe, 486 N.W.2d 407, 419 (Minn. 1992) (chairman of TWGDAM testified that “the guidelines were designed to be guidelines and not an inflexible set of rules”). 34/ The guidelines say These are only guidelines and should be used as a model for laboratory managers to set up an appropriate QA program for their laboratory. This document should not be construed as a mandate; it does not mean that failure to comply with each and every guideline, or that the use of an alternative or equivalent method is insufficient or likely to produce incorrect or unreliable results. Technical Working Group on DNA Analysis Methods, Guidelines for a Quality Assurance Program for DNA Analysis (April 1991) (“TWGDAM 54 Guidelines”) (Appendix II, Tab G). 55 Moreover, the 1995 TWGDAM guidelines have been replaced by the quality-assurance standards developed by the DNA Advisory Board (“DAB”), which became effective in October 1998. See DAB Quality Assurance Standards for Forensic DNA Testing Laboratories (1998) (“DAB Standards”) (Appendix II, Tab C).35/ See, e.g., Traylor, 656 N.W.2d at 899 (TWGDAM guidelines superseded by DAB standards); Cavin, Op. at 14 (Appendix I, Tab M)(same); Kirkendahl, Op. at 17-18 (Appendix I, Tab N)(same; also noting that even defense expert recognized the DAB as “an appropriate group to speak for the scientific community on appropriate guidelines”); see also Porter III, 1994 WL 742297, at *7 (recognizing that Congress authorized DAB to set policy for quality assurance in forensic DNA testing, and that TWGDAM guidelines would serve as interim standards until DAB standards were issued). Although the defendant argues that the TWGDAM guidelines should be followed because they have the 35/ Recognizing the importance of forensic DNA analysis, Congress enacted the DNA Identification Act of 1994, which authorized creation of a national convicted-offender DNA database, and established a DNA Advisory Board (“DAB”), composed of scientific, legal, and ethics authorities, to develop and recommend to the FBI quality assurance standards for forensic DNA laboratories. See Porter III, 1994 WL 742297, at *7. Members were appointed by the director of the FBI from a list of experts nominated by the National Academy of Sciences and professional criminal forensic societies. NRC II at 24. Using the 1995 TWGDAM guidelines as a model, the DAB developed new laboratory standards that were issued in final form in October 1998. The DAB expired in December 2000, and its oversight responsibilities reverted to TWGDAM, which is now called the Scientific Working Group on DNA Analysis Methods (“SWGDAM”). No changes have been made in the quality-assurance standards since the DAB expired. 56 imprimatur of the NRC, the NRC did not suggest that those were the only adequate quality-assurance guidelines. In fact, in its 1996 report, the NRC recommended that laboratories aspire to high quality standards “such as those defined by TWGDAM and the DNA Advisory Board.” NRC II at 88. “The DAB recommendations, which supersede TWGDAM guidelines, do not require that scientists developing new DNA technologies publish developmental validation studies in peer reviewed scientific journals.” Hill, 107 Cal. Rptr. 2d at 119; see DAB Standards, Appendix II, Tab C, at § 8. The DAB Standards do require that novel methodologies undergo “developmental validation to ensure accuracy, precision and reproducibility,” that such validation be documented, and that laboratories conduct internal validation studies before using new products. See DAB Standards, §§ 8.1.2, 8.1.3.36/ The kits and instruments used in this case met all of those standards. b. The Profiler Plus/Cofiler/310 Analyzer system has been proven reliable by laboratory validation studies, published peer-reviewed research, and actual experience. 36/ “Developmental validation” refers to the basic research, typically undertaken by a manufacturer, to ensure the reliability and accuracy of its testing products before selling them to laboratories. Laboratories perform “internal validation” studies, which the DAB standards define as “an accumulation of test data within the laboratory to demonstrate that established methods and procedures perform as expected in the laboratory.” DAB Standards (Appendix II, Tab C), Definitions, § 2(ff)(2). 57 In any event, whether or not the Perkin Elmer company publishes its developmental validation studies, the Profiler Plus/Cofiler/310 Analyzer system has been proven reliable in numerous other ways. The vast majority of courts agrees that DNA typing methodologies can be validated in a variety of ways, and that internal validation by the laboratories that actually use the methodologies is particularly persuasive. This makes sense. As one trial court that examined the Profiler, Cofiler, and 310 Analyzer system concluded, after considering the testimony of 11 expert witnesses: The system has been shown to work, time after time, by lab after lab, with or without studies from [Perkin Elmer]. The system is like a Model A Ford. Thousands of owners can tell us it works even if Henry Ford can’t or won’t explain it. The customers have thoroughly and scientifically validated this system. Dishmon, Order at 15 (Appendix I, Tab O). See also, e.g., Gaynor, Op. at 3 (Appendix I, Tab J)(quoting Dishmon with approval); State v. Traylor, 656 N.W.2d at 900 (reversing lower court, Minnasota Supreme Court held that PCR-STR testing methodology was generally accepted in the relevant scientific community; the standards of the DNA Advisory Board (“DAB”) were the appropriate standards in determining foundational reliability, as required by Minnesota law; the state laboratory complied with DAB standards; and admission of the DNA test results, without public disclosure of primer sequences or the manufacturer’s developmental validation studies, did not violate Traylor’s right to due process); Lynch, Ruling at 5 (Appendix I, Tab A) (relying on testimony 58 from court-appointed neutral expert witness, court concludes that, “Rather than focus on primer sequences[,] it seems that the focus should be on whether the results obtained from kits are reproducible. . . . Primer sequences are not necessary to see if the kits are working appropriately because labs can validate the results on their own.”); Elizarraras, Ruling at 5-6 (Appendix I, Tab C) (“lack of identification of the actual primer sequences is not fatal in being able to determine whether the [Profiler Plus] kit is working properly because the various laboratories can validate the results on their own.”); Bertsch, Tr. at 20377 (Appendix I, Tab F) (defense expert testified that he has no difficulty using Profiler Plus kit without access to the manufacturer’s validation information, including primers).37/ A Michigan court summarized the current legal consensus, saying that, “although developmental data has been released slowly and in an almost random manner, the extensive use of the system and recognized merits of . . . Perkin Elmer . . . testing kits show mutual national and even international 37/ Interestingly, it has been reported that a NIST scientist has determined the composition of the Perkin Elmer primers on his own, apparently through “reverse engineering.” See J. Butler and P. Vallone, “Determination of Commercial STR Kit Primer Sequences,” (abstract attached at Appendix II, Tab I, p. 105). See also Cavin, Op. at 39 (Appendix I, Tab M) (expert witness, an associate director of a paternity-testing laboratory, testifies that, “with access to proper lab,” he could determine Perkin Elmer primer sequences “in a relatively short time”). 59 acceptance in the scientific community.” Cavin, Op. at 47 (Appendix I, Tab M).38/ 38/ The one study cited by the defendant (Defendant’s Supplemental Motion at 20-21) does not support the argument that the primer sequences are flawed. See Shewale, J.G., et al., “Anomalous Amplification of the Amelogenin Locus Typed by AmpFLSTR Profiler Plus Amplification Kit,” 2(4) Forensic Science Communications (October 2000) (Appendix II, Tab Q). The study found "anomalous amplification" in 3 out of 7,600 samples at the amelogenin (sex typing) locus when Profiler Plus was used. As a result, the allele for the X chromosome did not amplify, although, because all 3 anomalies occurred in male samples, they did not lead to an inconclusive result as to the sex of the contributor. The study's authors concluded that the most probable explanation for the anomaly was the presence of a mutation at the primer binding site. The study does not suggest that Profiler Plus is not a reliable method of DNA typing and analysis, or that the primer used in that particular kit needs to be changed; rather it alerts analysts to the possibility of such an anomalous, but “rare” result and cautions that more careful interpretation of the data would be required when such an anomaly is observed. As a result, analysts will be able to recognize such an anomaly and take appropriate steps (such as retesting the sample). Scientific literature often may identify issues that affect the use of a scientific method, but that does not mean that the method itself has become unreliable. For example, phenomena such as potential contamination, are routinely addressed in laboratory procedural manuals such as the FBI’s “Short Tandem Repeat Analysis Protocol” (“FBI Protocols”) (Appendix II, Tab F). 60 Contrary to the defendant’s argument, validation studies have been made available to the scientific community. This Court should take judicial notice of the conclusion of the Michigan trial court in Kopp, which, after reviewing the testimony of several experts, as well as the scientific literature, held that the “[PCR/STR] methodology and the Perkin Elmer kits [Profiler Plus and Cofiler] as well as the 310 and 377 Genetic Analyzer have been subjected to a staggering amount of what can only be called peer review.” Kopp, Op. at 13 (Appendix I, Tab K). Perkin Elmer has described its developmental studies in numerous public documents and scientific forums; scientists from the FBI’s research lab have testified that they participated in the developmental studies with Perkin Elmer scientists; and, numerous laboratories, including the FBI DNAanalysis lab, have conducted internal validation studies comparing the results of the Perkin Elmer Profiler/Cofiler/310 Analyzer analyses with the results of older, previously validated systems, and found the results from the Perkin Elmer systems to be reliable.39/ Moreover, at least three more studies of the Profiler Plus/Cofiler and 310 Analyzer have been published in peer-reviewed scientific journals.40/ 39/ See, e.g., Lynch, at 5 (Appendix I, Tab A)(court-appointed neutral witness testifies that validation studies of Profiler and Cofiler kits have been done by scientists at Perkin Elmer, the FBI DNA lab, and the Virginia and Florida crime labs, all in accordance with the national quality-assurance standards promulgated by the DNA Advisory Board); Flores, 12/28/00 Tr. at 162-66 (Appendix I, Tab G)(summarizing scientific literature in support of reliability of Perkin Elmer kits); Kirkendahl, Op. at 18-20 (Appendix I, Tab N)(reviewing various ways in which Perkin Elmer systems had been 61 B. The general acceptance of DNA typing methods does not turn on whether the validated). 40/ Holt, C.L., et al., “TWGDAM Validation of AmpFlSTR PCR Amplification Kits for Forensic DNA Casework,” 47(1) Journal of Forensic Sciences 55 (2002) (Appendix II, Tab W) (conducting validation studies designed as recommended by the TWGDAM guidelines and DAB standards; studies showed that kits produced consistent and reliable results); Moretti, T., et al., “Validation of Short Tandem Repeats (STRs) for Forensic Usage: Performance Testing of Fluorescent Multiplex STR Systems and Analysis of Authentic and Simulated Forensic Samples,” 46(3) Journal of Forensic Sciences 647 (2001) (Appendix II, Tab U) (evaluating PCR/STR typing kits from Promega Corp. and Perkin Elmer; concluding that “commercially available multiplex kits can be used to amplify and type STR loci successfully” and that “[t]here was no evidence of false positive or false negative results”); Moretti, T., et al., “Validation of STR Typing by Capillary Electrophoresis,” 46(3) Journal of Forensic Sciences 647 (2001) (Appendix II, Tab V)(concluding that 310 Analyzer “provides efficient separation, resolution, sensitivity and precision, and the analytical software provides reliable genotyping of STR loci”). The Court should reject the defendant’s attempt to discredit research conducted by FBI scientists (Defendant’s Supplemental Motion Mot. at 22-23). FBI researchers are widely acknowledged to be leaders in the field of forensic DNA analysis. Their work in this area has withstood the scrutiny of scientists around the world for more than a decade. A desire by some FBI researchers to publish more of the results of their research is an understandable response to arguments, like that of the defendant, that scientific methods can only be validated by publication in peer-reviewed journals. The defendant also seems to misunderstand the peerreview process when he says that FBI examiners are not the “peers” envisioned by the courts. When FBI researchers submit an article to a peer-reviewed journal, it must pass an initial peer review by outside experts before the journal will agree even to publish the article. Upon publication, the research is open to even further scrutiny by other scientists. 62 laboratory applying those methods is subject to blind or open proficiency testing. The defendant’s complaint about the lack of blind proficiency testing at the FBI lab may go to weight, but it does not preclude the admissibility of the DNA evidence. See, e.g., Bridgett, 120 Daily Wash. Law Rptr. 155, at 1704 & n.29; Porter I, 1991 WL 319015, at * 21-22; see also, Sok, 683 N.E.2d at 681 n.21 (absence of blind proficiency testing not automatic bar to use of DNA evidence). The FBI laboratory follows the quality-assurance guidelines promulgated by the DAB, under which its analysts must undergo frequent open proficiency tests, and the practices of the laboratory have been certified by the Laboratory Accreditation Board of the American Society of Crime Laboratory Directors. Neither the superseded TWGDAM guidelines nor the most current recommendations of the NRC support the defendant’s assertion that the acceptance of a DNA-typing technique depends on whether the lab performing the technique has passed blind proficiency tests. The defendant’s reliance on TWGDAM rests on two fallacies: that the TWGDAM guidelines were intended to be mandatory standards for acceptance of DNA typing methods, and that the guidelines required, rather than expressed a preference for, blind proficiency tests over open proficiency tests (Defendant’s Supplemental Motion at 14-18). As previously shown, the TWGDAM guidelines do not impose mandatory standards for general scientific acceptance. Even if the guidelines 63 were conditions precedent to scientific acceptance, they do not say that blind proficiency tests are mandatory. See 1991 TWGDAM Guidelines (Appendix II, Tab G), § 9.2 (“It is highly desirable that the DNA laboratory participate in a blind proficiency test program, and every effort should be made to implement such a program.”); NRC II at 79 (“TWGDAM guidelines recommend one full-blind proficiency test per laboratory per year if such a program can be implemented”) (emphasis added). That blind proficiency testing may be "highly desirable" does not mean that its absence renders a laboratory's DNA-typing results inadmissible. Nor were the quality-assurance recommendations of the NRC intended to be an obstacle to findings of scientific acceptance. The thrust of the quality assurance section of the first NRC report was the urgent need – in 1992 – to develop reliable standards for forensic DNA typing. Thus, the report recommended development of national quality assurance standards, frequent proficiency testing of lab personnel, and monitoring and accreditation by external organizations. E.g., NRC I at 104-105. In a couple of isolated statements, the report endorsed blind proficiency testing as the ideal method of estimating lab error, without any substantive comparison of the benefits and drawbacks of blind versus open testing. See NRC I at 88, 106. See Bridgett, 120 Daily Wash. L. Rptr. 155, at 1704 n.33 (Court “is not compelled under the Frye standard to require the ideal but only the generally acceptable”). 64 The second NRC report clarified that the recommendations regarding quality-assurance standards were just that – policy recommendations – and strict adherence to those recommendations was not a prerequisite for scientific acceptance of lab results. See NRC II at 185.41/ In addition, the 1996 NRC committee found that blind proficiency tests, although seemingly advantageous in theory, are impractical, and, for that reason, should not be mandatory: [T]he logistics of constructing fully blind proficiency tests [to ensure that the forensics laboratory will not suspect that it is being tested] are formidable. The “evidence” samples have to be submitted through an investigative agency so as to mimic a real case, and unless that is done very convincingly, a laboratory might well suspect it is being tested. NRC II at 24.42/ 41/ The 1996 committee explained that, in the 1992 report, “[w]e emphasized the importance of minimizing laboratory error . . . . [However,] [t]hose recommendations rest not on a judgment that current error rates are so high that test results are scientifically unacceptable, but on a desire to reduce the incidence of errors to an extremely low value.” NRC II at 180. 42/ The second NRC committee had the benefit of the preliminary results of a Congressionally mandated study which concluded that a widespread policy of blind proficiency testing was unworkable. See “The United States DNA Advisory Board and Its Role in Setting Quality Assurance Standards,” Remarks by Arthur J. Eisenberg, 1st International DNA User’s Conference, November 24-26, 1999, at 2 (National Institute of Justice (“NIJ”) study concluded that “blind proficiency tests via law enforcement agencies, conduit laboratories and/or blind analyst models, is not feasible at this time”) (emphasis in original) (Appendix II, Tab N). The NIJ study did not, as the defendant mistakenly contends, endorse the use of blind proficiency tests (Defendant’s Supplemental Motion at 14). See NRC II at 79-80. The final 250-page NIJ report, entitled "Developing Criteria for Model External DNA Proficiency Testing," 65 The Porter courts anticipated the later conclusions of the NRC, when they rejected the argument that the admissibility of DNA evidence depended on a showing that the testing laboratory had complied with certain quality-assurance standards: [T]he defendant’s multi-prong arguments regarding proficiency testing are largely public policy pronouncements which have not been the basis of any court’s decision regarding the admissibility of scientific evidence. Simply put, if a new scientific evidence is generally accepted by the scientific community, whether the technique which generates it is performed competently is a matter of weight to be assigned by the jury. Porter I, 1991 WL 319015, at *22; accord Porter II, 618 A.2d at 636; see also Bridgett, 120 Wash. Law Rptr. 155, at 1703 n.27 (“[T]he Frye standard does not require the admission of utopic scientific methodology, but rather generally accepted scientific can be accessed at www.uic.edu/pharmacy/depts/forensicsci/. 66 methodology.”).43/ For the same reasons, this Court should reject the defendant’s complaints regarding blind proficiency testing.44/ 43/ Although Judge Kennedy acknowledged in Porter III that the first NRC report had endorsed the use of blind proficiency tests, he also recognized the difference between policy recommendations and factors that affect the scientific acceptance of new methods under the Frye test. He declined Dr. Eric Lander’s recommendation that the court decide whether a specific kind of proficiency testing program was appropriate, saying that “[t]he specific features of a proficiency testing program are better left to policymakers, i.e., legislators.” Porter III, 1994 WL 742297, *7. Accordingly, Judge Kennedy held that he would defer to the recommendations of the congressionally authorized DNA Advisory Board. Id. The DAB standards do not mandate blind proficiency testing. 44/ We address the defendant’s related argument – that the results of the proficiency tests should be combined with the probability statistics – in the section on random match probabilities, infra. 67 Even if it were necessary under Frye to demonstrate compliance with a sound quality assurance program, the FBI has a rigorous program of proficiency testing that complies fully with the current quality assurance standards issued by the DAB, which are consistent with the conclusions of the 1996 NRC report.45/ See FBI “DNA Analysis Unit I Quality Assurance Manual” (“Quality Assurance Manual”) (Appendix II, Tab C). The DAB standards say that “examiners and other [designated] personnel . . . shall undergo, at regular intervals of not to exceed 180 days, external proficiency testing . . . [which] shall be an open proficiency testing program.” See DAB Standards (Appendix II, Tab C), § 13.1. Consistent with those currently accepted quality control standards, each of the FBI DNA lab examiners must pass open proficiency tests at least every 180 days. See Quality Assurance Manual(Appendix II, Tab C), § 17.2.1 . To date, no examiner has called a match where there was none. In addition, since the Porter decisions, the FBI DNA analysis lab also has participated in the accreditation program of, and been accredited by, the Laboratory Accreditation Board of the American Society of Crime Laboratory Directors. See 45/ The defendant places undue emphasis on the references in the 1992 NRC report to an unusually high number of errors in two blind proficiency tests conducted at a private California laboratory (Defendant’s Supplemental Motion at 15-16). The 1996 NRC report, which had the benefit of several more years of experience with DNA analysis for forensic purposes, concluded that those errors were unlikely to be repeated because of increased emphasis on quality control. NRC II at 86. 68 Certificate of Accreditation, Appendix II, Tab D, and FBI analysts routinely testify about its accreditation status. Although the Melton “minimal reliability” standard is not implicated in this case, if it were,” the FBI’s decision not to submit to blind external proficiency testing does not undermine minimal reliability under Melton.” (Roberts/Veney Order at 18). We agree that, although the risk of error in a forensic analysis is not a factor affecting admissibility, it is a proper subject for a jury to consider when deciding the weight to be given the results of such analyses. For that reason, FBI experts routinely testify about the results of recent proficiency tests. That practice is consistent with what Judge Kennedy expected the DNA expert to do in Porter III, 1994 WL 742297, *7 n.22 (the government “is prepared to report that the FBI has made no false matches in its seven blind proficiency tests. This is acceptable, not misleading, and can be easily understood and evaluated by the jury.”); id. at *8 (error rate shall be stated by “presentation of expert testimony that the FBI laboratory has made X errors in Y blind proficiency tests”; also noting that such a presentation was in accordance with the [1992] NRC Report). It is also what Judge Richter contemplated in Bridgett. 120 Wash. L. Rptr 155, at 1703. C. The general acceptance of a scientific method does not turn on whether the quality assurance standards and protocols 69 for that method are also generally accepted, or whether the laboratory applying that method has followed accepted protocols. The defendant continues to quarrel with the fact that “[t]here is no basis in the law of this jurisdiction for holding hostage new scientific evidence which is “generally” accepted because of the absence of standards.” Porter I, 1991 WL 319015, *28 n.59. Yet they have produced no contrary authority in this jurisdiction, and, as shown in this opposition, the few cases from other jurisdictions that arguably hold otherwise are not persuasive. Although the defendant suggests that this Court does not have to follow Porter I, the trial court’s reasoning was adopted by the Court of Appeals. The Court of Appeals was well aware of the Porter trial court’s extensive discussion of the defense argument that the results of an accepted scientific method could not be admitted because the government did not show that the laboratory had followed generally accepted standards and protocols. See Porter I, 1991 WL 319015, *21*22; Porter II, 618 A.2d at 635 & n.9. After noting that the trial court had addressed this contention, and other contentions, “in considerable detail,” the Court of Appeals affirmed, “for the reasons stated by the trial court,” the trial 70 court’s conclusion that the RFLP method of DNA analysis was nonetheless generally accepted. Porter II, 618 A.2d at 635, 636.46/ If the RFLP methods could be found generally accepted in the relevant scientific community at a time when there were no industry-wide mandatory quality assurance standards, or wide-spread external accreditation programs, a fortiori, it is no barrier to a finding of general acceptance that the defendant thinks that the extensive quality assurance standards created by the DNA Advisory Board are insufficient. Even if the existence of sufficient standards were required for a finding of general acceptance, the defendant’s complaints come down to three issue: 1) the absence of public disclosure of the primer sequences used in Profiler Plus and Cofiler; 2) the alleged failure to ensure external validation of the Profiler Plus/Cofiler/310 Analyzer system, and 3) the 46/ One of the other contentions was the closely related contention that the DNA evidence could not be introduced because the laboratory had failed to perform the right kind of proficiency tests. The Porter I trial court reached the same conclusion about objections to the nature of proficiency tests, as it did about objections to lack of industry-wide standards. As with his objections, regarding the absence of uniform standards, the defendant’s multi-prong arguments regarding proficiency testing are largely public police pronouncements which have not been the basis of any court’s decision regarding the admissibility of scientific evidence. Simply put, if a new scientific evidence is generally accepted by the scientific community, whether the technique which generates it is performed competently is a matter of weight to be assigned by the jury. Porter I, 1991 WL 319015, *22; accord Porter II, 618 A.2d at 635-36. 71 fact that forensic laboratories are not required to undergo blind proficiency tests. The government has already demonstrated why those arguments do not undermine the general acceptance of the methods used in this case. The affidavit of Arthur Eisenberg, former chairperson of the DNA Advisory Board (DAB), explains that the TWGDAM guidelines, on which the defendant formerly relied for the position that revelation of primer sequences and evaluation of laboratory performance by blind proficiency testing were mandatory, did not intend to make those recommendations a mandatory requirement for scientific acceptance. Eisenberg Aff. ¶11. He also explains that the DAB modified those particular provisions because experience and research indicated that they were no longer advisable. Eisenberg Aff. ¶¶15, 23-28, 33-34. The DAB was created by Congress to create industry-wide standards for DNA forensic laboratories. Eisenberg Aff.¶12. It was recognized by the National Research Council and the Porter III court to be the appropriate body to create such standards. NRC II at 88; Porter III, 1994 WL 742297 *7 (D.C. Super. Ct. Nov. 17, 1994). Even if it were appropriate for this Court to determine the correctness of the quality assurance standards applied to forensic laboratories, the defendant has offered no reason why the Court should substitute the defendant’s view of sufficient standards for the judgment of the scientists appointed to the DAB (or those appointed to the NRC Committee on DNA 72 forensic science who believed that the DAB was an appropriate body to develop such standards). The FBI maintains detailed quality control and quality assurance procedures to regulate the performance of forensic PCR and STR typing in its laboratory. From a quality control standpoint, the FBI has developed a protocol for performing the PCR amplification and typing of the thirteen core STR loci using the Profiler Plus and Cofiler amplification kits. Budowle Tr. at A-155; App. II Tab F.47/ The protocol provides a procedure for performing each stage of the amplification and typing process, from extracting the samples through the typing stage. Id. The protocol was developed after research and validation studies of the technology and methodology by the FBI laboratory research unit, as well as after internal validation studies by the case working unit. Budowle Tr. at A-156. Since its development, the FBI has widely disseminated its STR protocol, including presentations at scientific meetings. Id. at A-157. There do not appear to be any peer-reviewed scientific publications that have found the FBI protocol to be unreliable or invalid for use in performing STR typing. Id. 47/ Citations to the expert testimony from the Trala case will be designated by witness name and page number as indicated in those original transcripts. The transcripts are bound in two appendices to the brief; Appendix IV, Part One, and Appendix IV, Part Two. The dates of the testimony are designated by a letter: May 7 - A; May 8 - B; May 9 - C; May 10 - D. 73 If one uses properly calibrated instruments and follows the FBI protocol, one should obtain essentially the same result (within established measurement error conditions) each time a technique is applied to the same DNA sample. Budowle Tr. at A-158. That is, while any analytic system will be subject to measurement error (as well as human errors), if the protocol is followed the results will have a zero error rate insofar as obtaining a wrong result (i.e., typing sample “B” as sample “A”) is concerned.48/ Id. at A-158-59. As for quality assurance, the FBI laboratory maintains a Quality Assurance Manual for the forensic case working unit, as well as an audit process. Budowle Tr. at A-159, App. II Tab E. Some of the specific procedures used to ensure quality assurance and quality control in the laboratory include: · · · · · maintaining separate rooms with dedicated equipment for pre- and post-amplification samples; use of hoods and safety sheeting with chemical operations; use of gloves, masks, and pipettes; separation of the extraction of the question and reference samples; use of positive and negative controls in the amplification and typing process;49/ 48/ Measurement errors are common to any analytic system. Take, for example, the measuring of a person’s height with a ruler. The same person could be measured at five-eight and one quarter on one occasion, and measured ay five-eight and three-eighths on another occasion. Although the results here are different by virtue of measurement error (assuming the person has not grown), the results can be considered the same within established tolerance limits for scientific purposes. Budowle Tr. at A-158. 49/ A positive control is a known sample of DNA that is run in every PCR reaction; they typing result of the positive control should be the same every time. Budowle Tr. at A-162. Negative controls generally contain no DNA in them. When these controls are 74 · technical peer review and administrative review of case work. Tr. at A-160-63; D-69. Thus, even if the Court were required to approve of admissibility standards before admitting DNA evidence, the standards followed by the FBI would pass muster. D. Allegations of interpretive errors in individual cases go to weight, not admissibility. 1. The Frye test does not require proof of adherence to correct procedures in each particular case. run in each reaction, they should show no result at the end of the process; if they generate results, it indicates the presence of a contaminant in the reaction. Id. at A-163. 75 It is not the law of this jurisdiction that "the proponent of the DNA [evidence] must also satisfy and demonstrate adherence to correct scientific procedures in the particular case at hand" (Defendant’s Supplemental Motion at 10).50/ The defendant’s reliance on cases from other jurisdictions which apply such a standard, such as United States v. Two Bulls, 918 F.2d 56 (1990), vacated on other grounds, 925 F.2d 1127 (8th Cir. 1991), and People v. Castro, 545 N.Y.S.2d (Super. Ct. 1989), is misplaced. The same case authority and argument was presented to Judge Kennedy in Porter I, but he refused to examine the reliability of the particular test results in Porter because neither the traditional Frye 50/ The defendant tries mistakenly to derive such a rule from United States v. Melton, 597 A.2d 892 (1991) (en banc) (Defendant’s Supplemental Motion at 7). Melton addressed whether a trial court must find the factual information on which an expert bases her opinion independently admissible and reliable before allowing testimony about the opinion. That case involved a psychiatrist's reliance on hearsay statements from hospital records and the defendant's family members to opine on the defendant's mental health. The scientific acceptance of the methods applied by the expert to those facts was not at issue, so Melton is not a gloss on the Frye standard. Nor does that case create an evidentiary obstacle to the admissibility of the DNA evidence in the instant case. Relying on the principles expressed in Federal Rule of Evidence 703, the Court of Appeals held in Melton that the facts or data on which an expert witness relies for her opinion do not have to be independently admissible, so long as they are "of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." Id. at 901. Because this was intended to be a “deferential” standard, id, the en banc Court overruled the panel's holding that the underlying information, if not otherwise admissible, must be "of a type for which the underlying reliability of the data can be sufficiently explored through cross-examination of the testifying expert." Id. at 904. Instead, the en banc Court required only that the 76 test nor precedent in this jurisdiction permitted such an examination.51/ Porter I, 1991 WL 319015, at *18 (holding that many of the defendant's objections to admissibility were "simply irrelevant, perhaps because of his erroneous equation of reliability with general acceptance"). The Court of Appeals expressly affirmed Judge Kennedy's understanding of the law in this jurisdiction, Porter II, 618 A.2d at 634, and emphasized that “[a]ny failure by the scientists to adhere to the appropriate procedure . . . does not raise an issue which implicates Frye.” Id. at 636. Given that express holding, and the Court of Appeals’ admonition in an earlier case that “[t]he answer to the question about the reliability of a scientific technique or process does not vary according to the circumstances of each case,” Jones, 548 A.2d at 40, the suggestion at the end of Porter II that it sometimes might be appropriate to conduct a pretrial inquiry into how well a procedure was performed must be read narrowly, and in the context of the state of the art of DNA analysis in 1992. See 618 A.2d at 644 n.28. underlying information be of minimal reliability. 51/ Judge Kennedy recognized that, even if a trial court agreed with the approach taken in Two Bulls and Castro, "trial judges in this jurisdiction are not permitted to disregard precedent because of an ad hoc determination that it is 'wrong.'" Porter I, 1991 WL 319015, *16 n.45. 77 The precaution of a pretrial "reliability" hearing on DNA typing might have been necessary on occasion in 1992, but such precautions are not necessary today. As Judge Richter has noted, “The world of forensic DNA science has not remained unchanged since DNA was first admitted in a criminal proceeding. If anything, DNA science has matured, becoming increasingly familiar, widely used, powerful, discriminating, and accurate.” (Roberts/Veney Order at 4-5). In 1992, DNA typing for forensic purposes was still relatively new; there was more variation in the products used to conduct the RFLP typing method; the forensic community had only begun to develop national quality-assurance standards for DNA analysis; and forensic laboratories were not routinely accredited in DNA procedures by external organizations. Examination of the application of accepted methods in a particular case is unnecessary twelve years later, especially in cases like the ones before the Court, in which the laboratory has applied a standardized and widely used PCR/STR system such as the Profiler/Cofiler/310 Analyzer system; the laboratory adheres to accepted national quality-assurance standards such as those promulgated by the DNA Advisory Board; and it is accredited by an outside organization such as the American Society of Crime Laboratory Directors Laboratory Accreditation Board.52/ See NRC II at 76-78 (components of high laboratory standards program 52/ The reasoning in Bridgett, 120 Daily Wash. Law Rptr. 155, at 1697, does not compel a different conclusion. After finding evidence derived from the RFLP process admissible under Frye, Judge 78 include adherence to quality-assurance guidelines and accreditation by professional organizations).53/ 2. The defendant’s allegation that the methodology applied in this particular case is unreliable does not demonstrate deficiencies in the underlying scientific methods. The defendant broadly claims that “The DNA examiner’s methodology in determining the profiles contained in each sample from the crime scene is unreliable and, therefore, renders his report of these profiles and his statistical conclusions inadmissible under Frye/Dyas.” (Defendant’s Supplemental Motion at 24). First, the defendant fails to articulate any examples of interpretive error Richter decided to conduct a separate pretrial hearing “to guarantee that the FBI followed its own procedures in this case.” Id. at 1703. The Court said that, in the absence of appellate guidance, it was taking this extra step solely as a prudential matter, and observed that “the precaution of a pretrial hearing should not be necessary, once admissibility is more firmly established.” Id. Bridgett was decided before the Court of Appeals reaffirmed, in Porter II, that how a laboratory performed the accepted scientific tests was a matter of weight, not admissibility. 53/ It is conceivable that, in a rare case, a laboratory might deviate from accepted methods so significantly that the reliability of the underlying methods is undermined. However, "[a]n allegation of failure to properly apply a scientific principle should provide the basis for exclusion of an expert opinion only if ‘a reliable methodology was so altered . . . as to skew the methodology itself.’” United States v. Martinez, 3 F.3d 1191, 1198 (8th Cir. 1993)) (citation omitted). The defendant’s complaints do not remotely rise to that level. 79 which could lead this court to conclude that the methodology used by this particular examiner was unreliable. Second, even if the defendant could do so, “[t]he mere fact that a body of data may be interpreted in two different ways does not establish a lack of general acceptance of the technique, experiments, or procedures by which those data were produced.” Whittey, Order at 14, 15 (Appendix I, Tab R) (holding, under Frye test, that such interpretative issues “pose no bar to the admissibility of PCR-based STR test results”). Challenges to an analyst's interpretation of data in a particular case are distinguishable from claims that a particular method or system cannot "repeatedly produce the same electropherogram from a sample of the same genetic material," id. at 15. The latter claims may affect admissibility under a Frye or Frye-Plus standard; the former do not. Further, to the extent that there is any argument by the defendant that interpretive errors occurred, he has not identified any expert who would support the idea that those mistakes undermine the scientific validity of the underlying methods, or even the weight that a factfinder should give the evidence as a whole. Because any allegations involving the application of methodology by a particular lab analyst raise no more than issues concerning the weight to be given the evidence, and not admissibility issues, they are not the proper subject of a Frye hearing. Porter II, 618 A.2d at 634, 636; see also Whittey, Order at 14, 15 (Appendix I, Tab R). 80 The defendant also argues that the FBI’s work does not pass muster under Frye because of problems with a particular analyst, Jacqueline Blake, and an investigation launched by the U.S. Department of Justice Inspector General (Defendant’s Supplemental Motion at 1 n.3). Ms Blake was a laboratory biology technician in one of the FBI Laboratory's two DNA Analysis Units who failed to perform a step in the standard operating protocol necessary for the detection of possible contamination that may have been introduced during analysis. She did not work on this case, and this problem does not extend beyond the work of that particular technician. Further, that technician's omission alone cannot cause a false positive result, and there is no evidence to date of a false positive among that technician's cases. 54/ Specifically, the technician violated the Laboratory's standard operating protocols by not processing negative control specimens to completion in approximately 100 cases over a two-year period. Those negative controls constitute a quality-assurance component of the analytical processing and reporting of DNA analysis by the method used in the instant case, which involves Polymerase Chain Reaction (PCR) amplification of Short Tandem Repeats (STR). There are two such negative controls: one is called a reagent blank, and it is added to the process at the initial extraction and purification step; the second is called a negative amplification control, and it is added at the step where the DNA obtained through extraction and purification is amplified for analysis. The negative controls function to monitor the PCR/STR process for the possible presence of contamination. The FBI Laboratory takes several measures to prevent contamination from being introduced during the PCR/STR process. The nature of the technician's omission affects the Laboratory's ability to confirm that contamination was not introduced during the DNA tests in which that omission occurred. However, even if contamination was introduced and went undetected in a particular test, it is highly unlikely that it could cause a false positive result. As stated in the scientific literature, contamination is more likely to cause inconclusive results, or a false 54/ 81 exclusion, than a false positive. The FBI Laboratory is taking steps to address cases handled by that technician, to include the retesting of evidentiary and known source materials where possible and appropriate. When retesting in the technician's cases has taken place and yielded viable results, those new results have not changed the ultimate conclusions regarding DNA profile inclusions or exclusions. 82 This matter was referred to the Justice Department's Office of the Inspector General and resulted in the resignation and subsequent conviction of hte technician. In addition, since learning of the technician's omissions, the Laboratory has revised its practices in an effort to detect and prevent such noncompliance with required protocols. Of greater importance, the analyst in the instant case has reviewed the underlying data, previously made available to the defendant in discovery, and can confirm that the negative controls were run in all of the tests in this case. Those controls confirmed that no contamination occurred. Thus, any omissions by Jacqueline Blake in unrelated cases clearly have little or no relevance to the analysis in this case. Nevertheless, to the extent relevant at all, any arguments the defendant may make about Jacqueline Blake or the investigation go to the weight of the evidence, not its admissibility. VI. The procedures by which the FBI determines statistical match probabilities, which include use of the product rule, are generally accepted by the relevant scientific community. Typing and comparison of DNA types at the selected loci are the first part of forensic DNA analysis. If a match is found, the analyst must next determine the significance of the match by calculating the coincidental, or random, match probabilities. The random match probability “essentially expresses the expected frequency of the observed DNA profile in a pertinent population.” United States v. Shea, 957 F. Supp. 331, 335 (D. Mass. 1997), aff’d, 159 F.3d 37 (1st 83 Cir. 1997). Our Court of Appeals has held that “the probability of a coincidental match is an essential part of the DNA evidence,” Porter II, 618 A.2d at 640, and, therefore, the statistical “procedures for determining a match must pass muster under Frye.” Id. at 631. The United States intends to offer expert testimony at trial regarding two types of DNA evidence that were recovered from the crime scene: (1) “sole source samples” that comprised blood from a single individual; and (2) “mixture samples” that contained blood from two or more individuals. With respect to the sole source samples, the government’s DNA expert will testify that the DNA profile of blood developed from thirteen sole source evidence samples matches defendant’s DNA profile, and that the DNA profile of blood developed from twenty two sole source evidence samples matches decedent’s DNA profile. Hence, there are a total of thirty five sole source samples. The government’s DNA expert will also discuss seven mixture samples. In some of the mixture samples, the FBI analyst will testify that the major contributor of sample is defendant and decedent can not be excluded as an additional contributor. Conversely, in other mixtures, the FBI analyst will testify that the major contributor is the decedent and the defendant can not be excluded as an additional contributor. Finally, in one mixture, the FBI analyst will testify that the decedent and defendant are both contributors. 84 Whether the evidence involves a single source sample or an indiscernible mixed sample, the procedure used by the FBI for calculating the random match probabilities begins by obtaining, from published tables, the frequencies of the alleles that were observed at each locus in the DNA profile from the examined sample. Shea, 957 F. Supp. at 335-36. Through studies of sample population databases, FBI and other research scientists have previously determined and published the frequency with which the alleles observed at each of the 13 CODIS loci have been found in the pertinent population. For a sole source DNA sample, or a mixed sample in which major and minor contributors can be discerned, or an intimate sample can be subtracted, the frequency of the genotype at each tested locus, i.e., the frequency with which the observed maternal and paternal alleles appear together at that locus, is calculated by established formulas based on a genetic principle known as the HardyWeinberg principle. For a mixed-sample of DNA – one in which there appear to be two or more contributors and there is no way to distinguish among the contributors – the FBI essentially adds the frequencies of all the possible combinations of alleles observed at a locus to obtain a frequency for the combined genotype at that locus. This approach is consistent with a statistical technique found acceptable by the NRC. Regardless of the method used to determine the frequency of a genotype at a particular locus, the frequencies of the genotypes at all the tested loci are multiplied together to obtain the 85 probability of a random match of the entire DNA profile. This last process is the product rule.55/ The FBI currently calculates match probabilities according to the methods approved by the National Research Council in its 1996 report (NRC II), which involve use of the product rule. See FBI Protocols, at pp. 10-6 to 10-8, Appendix II, Tab F. The defendant claims that these statistics cannot be admitted in evidence because the FBI does not incorporate a lab-error rate in its probability calculations; its use of the product rule is invalid because the databases from which its allele frequency charts are derived are too small, and use of the rule does not account for population substructuring; and the method used for mixed samples is not generally accepted in the relevant scientific community (Defendant’s Supplemental Motion at 24). The Court should reject these claims because the FBI’s methods for calculating random match probabilities are currently accepted by the relevant scientific community, as reflected in the recommendations of NRC II; that same scientific community has expressly rejected the defendant’s proposal to include lab-error rates in those calculations; and, as already shown, the alleged errors in interpreting data go to the weight of the evidence, not its admissibility. See Porter II, 618 A.2d at 636. 55/ The product rule provides that "if two events are independent of each other, the probabilities of each occurring can be multiplied, and the resulting product is the probability of both events occurring." State v. Link, 25 S.W.3d 136, 144 (Mo. Banc), cert. denied, 531 U.S. 1040 (2000). 86 A. Although not admissible twelve years ago, probability calculations based on the product rule are now admissible because the scientific consensus has changed. The defendant’s arguments about the size of the underlying databases and the existence of population substructuring were raised over twelve years ago in opposition to use of the product rule to calculate DNA match probabilities. Those concerns initially caused some scientists to hesitate to endorse that method, and that lack of consensus in turn caused our Court of Appeals to reject use of the unmodified product rule to calculate DNA match probabilities. See Porter II, 618 A.2d at 636. Scientific opinion has changed significantly since Porter II, and there is now consensus that the product rule can be used as the FBI now uses it to calculate reliable estimates of match probabilities. As Judge Richter has found, “Courts re-visiting the admissibility of DNA have found that that (sic) the scientific community no longer endorses the Porter II era safeguards, and accepts the product rule, either unmodified or with minor adjustments as described in NRC II.” (Roberts/Veney Order at 16). Thus, this Court can now admit estimates based on those methods. Reluctance to use the product rule arose primarily because of debate about the extent to which population “substructuring” existed, and the related extent to which substructuring could affect the reliability of match probabilities calculated by the product rule. See Porter II, 618 A.2d at 636-40; Bridgett, 120 Daily Wash. Law Rptr. 155, at 1704 & nn. 18-19; Porter I, 1991 WL 319015, at *2487 *28.56/ That perceived controversy caused the Porter trial court initially to refuse to admit match probabilities as calculated by the FBI at that time. Porter I, 1991 WL 319015, at *28. Because the court also held that the match statistics were an essential component of DNA profiling, it excluded the DNA evidence. Id. The government appealed. 56/ As Judge Kennedy explained, Substructure refers to distinct groups within a larger population whose members do not mate randomly with the rest of the subpopulation, but, rather tend to mate only with members of their group. If members of a subpopulation do not mate randomly with members of the larger population, that population may be out of HardyWeinberg equilibrium and the frequency with which certain alleles appear in the population could be different than the frequency of the same alleles in the subpopulation. Therefore, a database comprised of alleles from the larger population and which does not account for the allele frequency in the subpopulation could lead to an erroneous estimate of finding a coincidental match because the DNA of a suspect who is a member of the subgroup could be compared to the DNA of a population having genetic attributes different than his. Porter I, 1991 WL 319015, *10. 88 Between the trial court’s initial ruling and consideration by the Court of Appeals, the NRC published its first report on forensic DNA typing. That report acknowledged the debate over the use of the product rule to determine match probabilities. See NRC I at 80. Without resolving that debate, the report assumed that population substructuring may exist to a significant degree. Id. The NRC concluded, however, that the possibility that such problems existed should not prevent the introduction of DNA evidence because the assumed deficiencies “may be corrected by ensuring that the probability estimates of a coincidental match between the evidentiary and known samples are ‘appropriately conservative.’” Porter I, 618 A.2d at 643. NRC I proposed methods, called the “ceiling principle” and the “interim ceiling principle”, by which such conservative adjustments could be made.57/ NRC I at 80-85. Equating the NRC’s “conclusion regarding the proffer of a conservative method for calculating probability estimates” with “general acceptance of those methodologies in the relevant scientific community,” Porter II, 618 A.2d at 643 n.26, the Court of Appeals remanded the case back to the trial court to determine “whether the requisite consensus now exists for a conservative statement of the probability of a coincidental match and, if so, what the probability is in the present case.” Id. at 642-43. On remand, the trial court ruled 57/ For an explanation of how the ceiling principles were applied, see Bridgett, 120 Daily Wash. Law Rptr. 155, at *1704-05. 89 that probabilities derived from the modified ceiling principle recommended by NRC I would be admissible. Porter III, 1994 WL 742297, *5; see also Bridgett, 120 Daily Wash. Law Rptr. 155, at 1704-05. Those ceiling principles were acknowledged to place very conservative limits on the probability statistics. NRC I at 80. Since 1992, however, the scientific consensus has changed. The relevant scientific community no longer thinks it is necessary, or appropriate, to impose such conservative limits on statistics calculated with the product rule. This change has occurred in part because the PCR-based typing methods can define the alleles more accurately than the RFLP method, but mostly because studies have since shown that, to the extent there may be population substructuring that affects the frequencies with which particular alleles will be observed, the effect is more modest than originally hypothesized, and the effect can be compensated for in other ways. As a result, courts that have reexamined this question have concluded that the scientific community no longer endorses the ceiling principles, but, instead, accepts the product rule, either unmodified or with certain minor adjustments described in the 1996 NRC report, as an appropriate method by which to calculate match probabilities in forensic DNA analysis. See, e.g., People v. Soto, 981 P.2d 958, 976 (Cal. 1999) (collecting cases).58/ 58/ Among the recent decisions finding the unmodified product rule 90 There are at least three reasons why courts have concluded that scientists no longer think it is necessary to impose artificial ceilings when using the product rule, either for the older VNTR-based methods of analysis, like the RFLP method in Porter, or for the current PCR-based methods, like the STR method used in this case: to be generally accepted under Frye or Frye-Plus tests are: Soto, (Frye-Plus); Lindsey v. People, 892 P.2d 281 (Colo. 1995) (Frye), abrogated on other grounds, People v. Shreck, 22 P.2d 68 (Colo. 2001); People v. Miller, 670 N.E.2d 721 (Ill. 1996) (Frye); People v. Chandler, 536 N.W.2d 799 (Mich. App. 1995)(Frye); Watts v. State, 733 So. 2d 214 (Miss. 1999)(Frye-Plus); State v. Freeman, 571 N.W.2d 276 (Neb. 1997)(Frye-Plus); State v. Marcus, 683 A.2d 221 (N.J. Super. Ct. App. 1996) (Frye); Commonwealth v. Blasioli, 713 A.2d 1117 (Pa. 1998); State v. Copeland, 922 P.2d 1304 (Wash. 1996) (Frye); see also State v. Gore, 21 P.3d 262 (Wash. 2001) (product rule calculations admissible in cases involving PCR-based tests). Other cases have found general acceptance of the unmodified product rule as part of their relevance/reliability test under a Daubert-like standard. See, e.g., Commonwealth v. Rosier, 685 N.E.2d 739 (Mass. 1997)(finding product rule generally accepted for RFLP typing case); Taylor v. State, 889 P.2d 319 (Okla. 1995)(product rule passes State’s relevance/reliability test or Frye); see also Commonwealth v. Sok, 683 N.E.2d 671 (Mass. 1997)(relying on Rosier; product rule admissible in PCR case). 91 First, in 1993, the FBI conducted a survey of VNTR frequency data and determined that population frequency calculations based on the product rule were “reliable, valid and meaningful, without forensically significant consequences resulting from population substructure as had been postulated by some scientists.”59/ . . . [Second, b]ased on the FBI study, Dr. Eric Lander, the leading opponent in the scientific community to use of the product rule, declared that the “DNA fingerprinting wars are over”60/ . . . 59/ See Laboratory Division, Federal Bureau of Investigation, United States Department of Justice, VNTR Population Data: A Worldwide Study 2 (Feb. 1993) (“FBI VNTR study”). The conclusions of the VNTR study are summarized in Copeland, 922 P.2d at 1318-19. In addition to the FBI study, courts have relied on “extensive literature in peer-reviewed journals . . . [that] supports the premise that [population] substructuring does not impact significantly upon DNA population frequency estimates.” Soto, 981 P.2d at 976-77 (collecting cases); accord Blasioli, 713 A.2d at 1125-26; Copeland, 922 P.2d at 1319 (citing scientific journals). 60/ See E. Lander & B. Budowle, “DNA Fingerprinting Dispute Laid to Rest,” 371 Nature 735 (Oct. 27, 1994) (hereafter, “Lander & Budowle, ‘Dispute Laid to Rest’”) (see Appendix II, Tab J). Many courts found it significant that Dr. Eric Lander, a critic of the product rule who was one of the authors of the 1992 NRC report that recommended the ceiling methods, had changed his mind. In his 1994 article, Dr. Lander, and his co-author Dr. Bruce Budowle, said that NRC I “failed to state clearly enough that the ceiling principle was intended as an ultra-conservative calculation, which did not bar experts from providing their own ‘best estimates’ based on the product rule.” Lander & Budowle, at 737. They also concluded that the FBI’s population surveys yielded a remarkable database for examining allele frequency variation among ethnic groups. Reassuringly, the observed variation is modest for the loci used in forensic analysis and random matches are quite rare, supporting the notion that the FBI’s implementation of the product rule is a reasonable best estimate. Lander & Budowle, at 738. 92 Finally, in its 1996 Report [NRC II], . . . , the NRC noted “‘[t]he ceiling principles were intended for VNTRs with many alleles, no one of which has a very high frequency [and the ceiling principles] are not applicable to PCR-based systems.’” The 1996 Report concluded that both the ceiling and the interim ceiling principles were unnecessary and “‘[i]n general, the calculation of a profile frequency should be made with the product rule,’” both for VNTR and PCR-based systems. Watts v. State, 733 So. 2d 214, 225-26 (Miss. 1999)(internal citations omitted). See NRC II at 5. Accord Soto, 981 P.2d at 975-76 (Frye-Plus; RFLP); Commonwealth v. Rosier, 685 N.E.2d 739, 745 (Mass. 1997)(modified Daubert test; RFLP; based on same authorities, concluding that “the controversy has been resolved in large part”); State v. Freeman, 571 N.W.2d 276, 291-93 (Neb. 1997) (Frye test; RFLP; relying on Lander & Budowle article, NRC II, and expert testimony); Commonwealth v. Blasioli, 713 A.2d 1117, 1125-26 (Pa. 1998)(Frye; RFLP; relying on FBI VNTR study, Lander & Budowle article, NRC II); State v. Copeland, 922 P.2d 1304, 1318-1319 (Wash. 1996)(Frye; RFLP; relying on FBI VNTR study, Lander & Budowle article, and pre-publication copy of NRC II; “use of the product rule in establishing statistical probabilities of a genetic profile frequency in the human population is generally accepted within the relevant scientific community and . . . a significant dispute no longer exists on this matter”); cf. United States v. Lowe, 954 F. Supp. 401, 418-19 (D. Mass. 1997) (Daubert test; upholding admissibility of probabilities based on product rule where there was statistical independence among the loci examined by PCR-based processes (DQ-Alpha, Polymarker, and 93 D1S80)). Thus, the concerns that may have precluded scientific consensus over use of the product rule when Porter II was decided were resolved shortly thereafter, and it is no longer necessary to apply the ceiling principle when calculating random match probabilities.61/ Although the Court of Appeals in Porter II did not find a scientific consensus in favor of the unmodified product rule in 1992, that conclusion does not prevent this Court from deciding whether, in view of the advances in scientific knowledge since 1992, the product rule has now achieved scientific acceptance. The DCCA did not rule as a matter of law that the ceiling principles must be used indefinitely. The Court merely held that, when Porter II was decided, there did not yet appear to be a scientific consensus regarding use of the product rule for forensic DNA analysis purposes and, therefore, as required by Frye, evidence calculated using that rule was not yet admissible. 61/ As a result of the NRC’s conclusions in 1996, even the few remaining critics of the product rule “expressly concede [that] the product rule/population substructure issue has indeed been laid to rest.” Soto, 981 P.2d at 976. The Soto court was referring to concessions by two law professors who often testify as experts in opposition to DNA evidence: William Thompson and Richard Lempert. See Thompson, “Accepting Lower Standards: The National Research Council’s Second Report on Forensic DNA Evidence,” 37 Jurimetrics 405, 423 (noting that population structure studies have “tipped the balance of scientific opinion in favor of the product rule (or something close to it)”); Lempert, “After the DNA Wars: Skirmishing With NRC II,” 37 Jurimetrics 439, 455 (lack of HardyWeinberg and linkage equilibria “seems hardly to matter. Empirical studies suggest that conservatism in estimating allele frequencies in the first instance can more than make up for any prejudice an accused suffers from the untenability of the assumptions.”). 94 Porter II, 618 A.2d at 631. It would be contrary to the intent of Frye, and to the Court of Appeals’ express reluctance to withhold relevant information from the jury, see 618 A.2d at 640-42, to interpret the 1992 Porter decision as freezing the law in perpetuity. Other courts that initially rejected the product rule under a Frye or FryePlus standard have since concluded, first, that they are not barred by stare decisis from deciding whether the product rule is now an acceptable method by which to calculate match probabilities, and, second, that the scientific community has now accepted use of the product rule for that purpose. See Freeman, 571 N.W.2d at 293 (Frye-Plus jurisdiction; overruling prior decision that, relying on NRC I, required use of ceiling principles; “[S]cientific opinion on a particular scientific matter is not static, and therefore, a determination of whether a particular type of novel scientific evidence is or is not generally accepted within the scientific community must consider the current state of the science.”); Dishmon, Order at 18-19 (Appendix I, Tab O)(Frye-Plus jurisdiction; concluding that prior state supreme court decision approving use of interim ceiling method was not stare decisis; 1996 NRC II report finding interim ceiling method unnecessary and approving use of product rule “represents the best scientific thinking of the scientific community on the subject,” and thus product rule meets the Frye test as applied in Minnesota); accord, Kirkendahl, Order at 34-35 (Appendix I, Tab N)(Minnesota Frye-Plus test; 1996 NRC II report represents 95 current consensus among scientists). For the same reasons, this Court should conclude that match probabilities calculated with the product rule are now admissible. B. The FBI calculates probability statistics for single-source evidence samples by methods found acceptable by the National Research Council. The Court should admit the probability statistics developed by the FBI analyst pertaining to the thirteen sole source evidence samples that match the DNA profile of the defendant and the twenty two sole source evidence samples that match the DNA profile of the decedent. The formula the analyst applied is generally accepted as a reliable formula to apply to a single source profile. In place of the ceiling principles, “‘[t]he [NRC] now recommends the use of a modified version of the product rule which assumes the existence of some undetected population substructure of a lesser magnitude than that reflected by use of the ceiling principle.’” Blasioli, 713 A.2d at 1125 (internal citation omitted). The minor modifications recommended by NRC II appear in Recommendation 4.1. See NRC II at 5. As stated in its protocols for STR typing, the FBI routinely applies the modifications recommended by NRC II when it calculates match probabilities for single-source or major contributor samples. See FBI Protocols (Appendix II, Tab F), at pp. 10-6 to 10-8. 96 The FBI uses the modified random match probability (RMP) formula specified in NRC Recommendation 4.1 for single-source DNA profiles. See NRC II (App. II, Tab A) at 122. As noted, the basic methodology in that formula is premised on the product rule. Id. at A-178. In essence, the statistical frequency of the DNA profile at issue is calculated by multiplying the frequency of each of the alleles in the profile, and then correcting the result to account for inbreeding or substructuring effects in the population, as well as applying additional statistical qualifications. Id. at A-178; Chakraborty Tr. at D-104-05. The profile frequency that emerges from these corrections/qualifications is more conservative than would be obtained by use of the strict product rule. Chakraborty Tr. at D-106, 108. The NRC/FBI formula includes several conservative features to address substructuring or inbreeding. The first is the inclusion of a value denoted by the variable “theta.” Budowle Tr. at A-176; Conneally Aff. ¶8A. The degree of inbreeding in a given population can be calculated by empirical population studies. Budowle Tr. at A-167. As a result of such research, the NRC recommended a theta of .01 for all but the most isolated populations (in which case a value of .03 is recommended). See NRC II at 122.62/ The FBI laboratory 62/ The method of statistical analysis described above is not new to STR analysis, and was used in essentially the same manner with RFLP analysis of VNTRs. Budowle Tr. at A-186-87. Indeed, the concept of the theta inbreeding coefficient correction has been in use since the 1950s. Id. at A-187. 97 follows this recommendation, despite the fact that the weight of the current empirical research on various world populations suggests that the actual observed value of theta is well below .01, rendering the NRC II recommendation highly conservative. Id. at A-176-77, Chakraborty Tr. at D-109-114; Conneally ¶8A.63/ In addition, the NRC/FBI formula does not use the actual allelic frequencies for certain alleles that are very uncommon in the database. Chakraborty Tr. at D-106; Conneally Aff. ¶8A. In other words, if the actual frequency for a particular allele is below a particular threshold, the formula requires the analyst to assign that allele a higher, and thus, a more conservative, frequency for purposes of the calculations. Id. The FBI’s allelic frequency tables incorporate this “floor.” Conneally Aff. ¶ 8A. 63/ The defendant’s expert in Trala, Dr. Shields, advocates use of a theta value of .05, but his is a minority view that was first presented to the 1992 NRC committee. Tr. at D-26-27. The 1996 NRC report saw no need to adopt such a theta value. Tr. D-27-28. 98 The NRC also recommended that a tenfold tolerance limit be applied to statistical estimates based on the RMP formula. NRC II at 27; Conneally Aff. ¶8A. Application of such a “factor-of-ten” adjustment to an RMP of 1 in 100,000 would result in a margin of uncertainty from 1 in 100,000 to 1 in 10,ooo,ooo. Conneally Aff. ¶8A. This adjustment corrects for genetic or sampling variation that might occur. Conneally Aff. ¶8A. It generally conveys to the jury a more conservative result (i.e., one that overstates the actual frequency). Budowle Tr. at A-185. In the past, the FBI routinely applied that limit to its RMP estimates. Id.at A-185. That practice is apparently no longer routine; however, on direct examination, the government will ask the FBI analyst to apply that factor of ten adjustment, after he reports the raw probability estimates.64/ Because the FBI applies the NRC’s recommended formula for calculating random match probability estimates for single-source DNA profiles, and the NRC’s views are the equivalent of general acceptance, the Court should admit the sole source sample probability estimates. 64/ Even if the prosecutor forgets to ask the analyst to apply that adjustment, the analyst will give the same answer on crossexamination. 99 The modifications to the random match probability formula contained in NRC Recommendation 4.1 are appropriate if the underlying databases appear to be in Hardy-Weinberg and linkage equilibrium, or the deviations from HardyWeinberg or linkage equilibrium are not statistically significant.65/ The FBI typically reports random match probabilities derived from four databases that represent the major population groups in the United States: Caucasian, Black, Southeastern Hispanic, and Southwestern Hispanic.66/ Each of these major databases contains approximately 200 samples.67/ The frequency of the occurrence in each database of each allele at each tested locus has been 65/ “Hardy-Weinberg equilibrium . . . means that one allele at a locus is not predictive of the other allele at that locus . . . Hardy-Weinberg equilibrium depends upon an assumption of a large population in which there is random mating.” Copeland, 922 P.2d at 1317 (internal citations omitted). “[L]inkage equilibrium . . . means that the alleles at different loci are inherited independent of each other.” Id. 66/ Analysts will refer to other databases if the circumstances of a case suggest that a population subgroup (e.g., Apache, Vietnamese, Caribbean) should be considered in addition to the major groups. 67/ "These [STR] databases contain more than a statistically sufficient number of samples to permit the calculations of valid allelic frequencies." United States v. Trala, 162 F. Supp. 2d 336, 351 n.15 (D. Del. 2001) (noting also that FBI has published up to 41 of these databases and made the underlying raw data publically available). See Budowle, B., et al., “CODIS STR Loci Data from 41 Sample Populations,” 46(3) Journal of Forensic Sciences 453 (2001) (Appendix II, Tab T). See also State v. Salmon, ___ S.W.3d ___, 2002 WL 31245217, *5 (Mo. App. 2002) (Dr. Kenneth Kidd, professor of genetics at the Yale School of Medicine, testified that "it was a generally accepted practice in the forensic science community to use the FBI database" to calculate random match probabilities). 100 compared with worldwide data compiled by the FBI and with data from other laboratories’ databases. These studies amply demonstrate the acceptability of the use of the FBI's databases to determine allele frequencies, and the use of the product rule to calculate match probabilities from those frequencies.68/ These databases contain a more than statistically sufficient number of samples to permit the calculations of valid allelic frequencies at the thirteen core STR loci. Chakraborty Tr. at D-129, 166; Conneally Aff. ¶11.69/ 68/ See, e.g., Holt, C., et al., “Practical applications of genotypic surveys for forensic STR testing,” 112 Forensic Science International 91 (2000)(concluding that reliable multilocus profile estimates can be obtained by applying product rule to genotype frequencies at the 13 CODIS loci) (Appendix II, Tab R); Budowle, B., et al., “Population Data on the Thirteen CODIS Core Short Tandem Repeat Loci in African Americans, U.S. Caucasians, Hispanics, Bahamians, Jamaicans, and Trinidadians,” 44(6) Journal of Forensic Sciences 1285 (1999)(concluding that the application of the product rule is valid for estimating the rarity of a multiple loci profile for the 13 CODIS loci) (Appendix II, Tab M); Lins, A.M., et al., “Development and Population Study of an Eight-Locus Short Tandem Repeat (STR) Multiplex System,” 43(6) Journal of Forensic Sciences 1168 (1998)(concluding that the application of the product rule is appropriate for estimating the rarity of a PowerPlex profile, which examines eight of the loci examined by Profiler Plus)(Appendix II, Tab L); Budowle, B., et al., “United States Population Data on the Multiplex Short Tandem Repeat Loci – HUMTHO1, TPOX, and CSF1PO – and the Variable Number Tandem Repeat Locus D1S80,” 42(5) Journal of Forensic Sciences 846 (1997)(concluding that the application of the product rule is appropriate for estimating the rarity of a multiple loci profile for these four loci, three of which are incorporated in Profiler Plus) (Appendix II, Tab K);. 69/ The fact that the databases are convenience samples rather that true random samples does not affect their validity. See Conneally Aff. ¶10; NRC II 30, 126-27. 101 The FBI has published up to 41 of these databases, and has placed all the underlying raw data on the Internet for public review. Budowle Tr. at A-172-73; App. II, Tab T. Since the time that this data was placed on the Internet, the FBI is unaware of any peer-reviewed scientific articles critical of the databases or suggesting that the databases are insufficient. Budowle Tr. at A-173; see also Conneally Aff. ¶15. Empirical research in peer-reviewed scientific journals has demonstrated that the population data used by the FBI to calculate allelic frequencies meets the statistical expectations - Hardy-Weinberg and linkage equilibrium underlying the formulas recommended by NRC II to generate a final profile frequency.70/ See Conneally Aff. ¶10. In particular, a recent study in the peerreviewed scientific journal Electrophoresis demonstrates that the thirteen core loci tested in this case could even be multiplied together using the strict (i.e., unmodified) product rule with statistically reliable results. Chakraborty Tr. at D116. 70/ Empirical research has demonstrated that there are sporadic departures from the statistical assumptions of independence underlying the use of the modified product rule for specific pairs of loci in specific databases. Chakraborty Tr. at D-124. However, the number of occurrences of such deviations is below the nominal level of statistical significance. Id. at D-123-24. As a result, even with these sporadic deviations, the thirteen core loci are sufficiently independent to allow for the use of the modified product rule. Id. at D-125. 102 The defendant contends that the FBI’s databases are improperly structured along racial lines, and that this is an ineffective way to account for population substructuring (Defendant’s Supplemental Motion at 25-26). The FBI does not rely on its database categories to compensate for possible substructuring. As shown above, the RMP formula used by the FBI contains its own provisions to account for possible substructuring. The FBI maintains separate databases for different population groups because empirical research still suggests that, for purposes of forensic identification by DNA analysis, there is some observable difference in allelic frequency rates between populations, although it is widely understood that the greatest degree of genetic difference will be lower within large population groups. Conneally ¶12. A simple examination of the FBI’s allelic frequency charts, App. II, Tab F, at 10-9 to 10-24, and the results of the probability calculations in this case reveals that there are still observable differences between larger populations. Nothing the defendant has presented demonstrates that the probability formulas endorsed by the NRC are no longer generally accepted in the relevant scientific community, or that those formulas should not have been used in these cases. The Court should therefore admit the evidence and the accompanying probability statistics. Moreover, the defendant’s suggestion that the interpretation of mixed samples is “mired in controversy” (Defendant’s Supplemental Motion at 26-27) is 103 unfounded. He relies solely on an abstract that discusses some of the issues that arise in the interpretation of DNA from mixed samples, that that reliance is misplaced (see Defendant Supplemental Motion at 26-27). The cited article describes several ways in which such complex mixtures can be assessed, but it nowhere suggests that any of the approaches that it describes is scientifically unacceptable. See Ladd, C., et al., “Interpretation of Complex Forensic DNA Mixtures,” 42(3) Croatian Medical Journal 244 (2001) (Appendix II, Tab S); see also Testimony of Frederick Bieber in State v. Roman Nose (Government Appendix VI, Tab B)(Dr. Bieber, a co-author of the Croatian Medical Journal article, explains that FBI methods are generally accepted).71/ Moreover, the 71/ Dr. Bieber is an Associate Professor of Pathology at Harvard Medical School who teaches mdical students, graduate students, post-doctoral fellows, and medical residents and interns, and directs two hospital laboratories that perform diagnostic genetic testing. He holds a Ph.D. in human genetics, and completed a postdoctoral fellowship in medical genetics. Among other pforessional affiliations, he is a member of the American Society of Human Genetics, the American Board of Medical Genetics, and the American Academy of Forensic Sciences, and he was appointed to the Congressionally-mandated DNA Advisory Board (“DAB”). Of particular significance to this case, Dr. Bieber was a co-author of the article “Interpretation of Complex Forensic DNA Mixtures” on which the defendant relies for the proposition that there is no consensus on an acceptable method by which to interpret mixed DNA samples. Dr. Bieber explains in his testimony that the challenges involved in interpreting mixed samples do not undermine the reliability of DNA typing by PCR/STR methods, or evidence a general inability on the part of DNA analysts to accurately interpret mixed samples, including, in appropriate circumstances, the ability to discern individual DNA profiles from mixed DNA samples as was done in one of the samples in this case (see, e.g., 10/28/02 Tr. 122-136). Dr. Bieber also testified that the random match probability formulas 104 DNA Advisory Board, which represented a cross-section of experts in the field of DNA typing, concluded that two of the methods discussed in the Ladd article – the "combined probability of exclusion" (which is a different way of stating the probability of inclusion figure calculated by the FBI), and the likelihood ratio were equally accepted ways of expressing the probative value of DNA evidence. See DNA Advisory Board, “Statistical and Population Genetics Issues Affecting the Evaluation of the Frequency of Occurrence of DNA Profiles Calculated From Pertinent Population Database(s),” 2(3) Forensic Science Communications (July 2000) (Appendix II, Tab O). In addition, because of the very low match probability estimates in this case, the Court should permit the analyst to testify to his opinion that, to a reasonable degree of scientific certainty, the defendant is the source of the DNA found on thirteen sole source evidence samples and that the decedent is the source of the DNA found on thirteen sole source evidence samples. Under certain statistical circumstances, there is a point at which it would be scientifically reasonable to attribute the source of a given DNA sample to an recommended by the NRC II report were generally accepted methods for calculating match probabilities for single-source DNA samples, or individual DNA profiles derived from some mixed DNA samples; that either probabilities of exclusion or likelihood ratios were acceptable methods for expressing match probabilities for mixed samples; and that error rates did not have to be included with probability statistics for those statistics to be acceptable (10/28/02 Tr. 62-66, 137-166, 168-171). 105 individual. NRC II at 136-37; Budowle Tr. at A-193. In particular, given a confidence level of 99 percent, and given the current United States population of 280 million, one can reach a source attribution conclusion to a reasonable degree of scientific certainty if the profile frequency of the ostensible source and the matching unknown sample is smaller than one in 280 billion. Tr. at A-195. The FBI has adopted this formula for reaching source attribution conclusions. Id. The probability statistics for all thirty five sole sourse evidence samples (13 matching the defendant; and 22 matching the decedent) in this case met the source attribution threshold. The NRC II committee concluded that source attribution was appropriate when the RMP was low enough. It elected, nonetheless, not to recommend any particular approach to determining the precise point at which the RMP would be low enough to permit source attribution, leaving that issue to the courts. NRC II at 136-37. It did, however, describe two methods for setting thresholds that would allow an analyst to make such an attribution. When the DAB reconsidered the issue, it concluded that, because DNA typing methods were able to type such a high number of genetic markers, the magnitude of random match probability estimates "has approached the point where it is unlikely that two unrelated individuals carry the same type." Budowle, et al., Source Attribution of a Forensic DNA Profile, App. II, Tab P. The DAB sought to develop a 106 formula to act as an objective measure by which an analyst could determine when it was appropriate to make such an attribution. Id. Although there has been debate over the DAB's recommendations, that debate has not been over whether it ever can be appropriate to attribute a source, but what is the best formulation for deciding when to reach such a conclusion.72/ The FBI’s formulation is generally more conservative than other approaches presented in the scientific literature. Budowle Tr. at C-36. In fact, the FBI’s source attribution formula has been described by Dr. James Crow, chair of the National Commission on the Future of DNA Evidence Research and Development Working Group, and a leading expert in the field of genetics, as “a very sensible one.” Tr. at D-10,11. As a result of the conservative approach taken by the FBI, Dr. Crow further concluded: So I think almost anybody would agree that [the FBI source attribution formula] is a procedure that is safe in the sense it would be very, very rarely would you make the mistake in the wrong direction; that is, in the direction of wrongly convicting an innocent person. . . . And I guess we’ll have to ultimately take the position, I would myself at least, that the FBI’s procedure is the right thing to do. I’m happy to say that the DAB [DNA Advisory Board] said the same thing. 72/ Even the defendant’s expert in the Trala case agreed that one can make a source attribution under appropriate circumstances. Shields Tr. at D-22. 107 Tr. at D-12-13 (emphasis added). The Commission’s final report, moreover, states that a “high degree of confidence” can be obtained in these source attribution conclusions. Tr. at C-106. The DAB adopted a source attribution formula much like one recommended by the NRC as valid for that purpose. Like the NRC, the DAB represented a cross-section of the relevant scientific community. Its views on the source attribution issue may fairly be equated with the views of the relevant scientific community. Accordingly, the Court should allow the analyst in this case to report a source attribution. C. The FBI’s method for calculating combined probabilities for mixed sample DNA profiles is generally accepted. When the analyst calculated the match probabilities for INSERT # of the mixed evidence samples as though he had no information about possible contributors, he used a formula known as the combined probability of inclusion (“CPI”). The CPI formula is inherently very conservative because it does not consider how many contributors are present in the unknown DNA samples based on other reliable information. Conneally Aff. ¶8B. Instead, the formula creates a frequency profile for each locus based on all the possible combinations of all the alleles actually observed at that locus. Whether there were two contributors or a dozen, makes absolutely no difference to the result. This ordinarily causes the combined frequency at each locus to be much higher 108 and, thus, much more favorable to a defendant than it would be if the frequency were determined by the RMP formula appropriately used for single source or major contributor profiles, which essentially multiplies alleles together. Whether the CPI or RMP formula is used to determine the frequency at each locus, after the combined allele frequency is determined for each of the tested loci, the product rule is used to calculate the frequency of the entire DNA profile, that is, the combined allele frequencies for each of the tested loci are multiplied together to determine the match probability for the entire DNA profile. The 1992 NRC committee expressly advocated the use of the combined probability of exclusion (CPE) formula (NRC I at 59), which is based on the same mathematical principle as the combined probability of inclusion (CPI) formula used by the FBI. The 1996 NRC committee preferred a different approach – the likelihood ratio – but it did not conclude that the CPE/CPI was an invalid method. NRC II at 129-130. Instead, the NRC II report said that the CPE/CPI approach was “hard to justify because it does not make use of some of the information available,” i.e., the genotype of the suspect. NRC at 130. The scientific community had another opportunity to examine the formulas that could be used to apply to mixed samples through the work of the DNA Advisory Board (DAB). Although defendant disparages the DAB, suggesting that it was a pawn for the FBI, that suggestion has no basis in fact. Dr. Arthur Eisenberg, who was the second chairman of the DAB, explains that, 109 although the Director of the FBI made the final selection of members, Congress ensured that the Board represented a broad and independent cross-section of the relevant scientific community by specifying the fields from which the members were selected, and by limiting the Director’s choice of members to persons nominated by the National Academy of Sciences, and professional organizations. Eisenberg Aff. ¶14. DAB members were wholly independent from the FBI. Eisenberg Aff. ¶14. The only control the Director had over the final recommendations of the DAB was his ability to veto them; but he did not do so. Eisenberg Aff. ¶¶14, 16. Because the membership of the DAB was diverse, and it actively solicited the views of the scientific community, Eisenberg Aff. ¶14, it is reasonable to treat its recommendations as reflecting a consensus within the scientific community familiar with DNA forensic analysis. The DAB affirmed that both the CPE/CPI and LR methods were acceptable methods to apply to mixed samples. See App. II, Tab O, at 5. It pointed out that the PE calculation “is particularly useful in complex mixtures, because it requires no assumptions about the identity or number of contributors.” Id. As Dr. Conneally observes, the choice of method depends more on policy considerations than scientific validity, because both formulas are accepted. Conneally ¶8B. The policy advantage of the CPE/CPI formula is that it is much simpler to calculate, and to express correctly to a jury. Id. The authors of the survey article about interpretation of complex mixtures, on which defendants 110 rely for many of their arguments, speak favorably of the CPE/CPI approach, saying that it “provides a straightforward, conservative, yet still highly informative calculation that avoids potential pitfalls associated with extrapolating the genotypes of contributors.” See App. II, Tab S, at 245. In contrast, the authors conclude that “[S]ome LR calculations and interpretations can be complicated, and their significance to the case may not be apparent to the practitioner and the trier of fact.” Id. Based on the acceptance of the CPE/CPI formula by the NRC committees, its endorsement by the DAB, and the inherently conservative nature of the CPE formula, this Court should admit the CPI probability statistics calculated by the analyst in this case. D. Inclusion of error rates in probability statistics is unacceptable in the relevant scientific community, and therefore not required under Frye. The court should reject the defendant’s contention that “the FBI must disclose its error rate information or limit its proffered random match probability by the effect of an estimated error rate using the confidence interval” (Defendant’s Supplemental Motion at 20). This is another issue that is not properly before the Court because, in the District of Columbia, laboratory error is an issue going to the weight of scientific evidence, not to its admissibility, see 111 Porter II, 618 A.2d at 636, a distinction that the defendant repeatedly ignores. Accordingly, this argument should be rejected without further consideration. Even if lab error were a proper subject for a Frye inquiry, the defendant’s proposals must be rejected because they are contrary to scientific opinion: “General acceptance does not require including error rate in probability calculations. The scientific community has rejected using error rates to temper the probability calculations.” (Roberts/Veney Order at 17). The NRC, whose views are treated by our courts as representative of scientific consensus on DNA issues, has expressly rejected the notion that error rates should be estimated from the results of periodic proficiency testing, or that estimates of error rates should be used to limit the random match probability numbers. The first NRC report concluded that “[c]oincidental identity and laboratory error are different phenomena, so the two cannot and should not be combined in a single estimate.” NRC I at 88. The second NRC report also explicitly rejected the idea that laboratory error rate was an appropriate component of the match probability calculations, or even that something approaching a rate of error could be reliably determined: The committee offers these recommendations to improve laboratory performance rather than to try to estimate the probability that a particular laboratory makes a mistake by reporting that two DNA profiles match when in fact they do not match. Auditing and proficiency testing cannot be expected to give a meaningful estimate of the probability that a particular laboratory has made such an error in a specific case. An unrealistically large number of proficiency tests would be needed 112 to estimate accurately even an historical error rate. For such reasons, proficiency test results should not be combined with the estimated frequency of an incriminating profile to yield the probability that a laboratory would report that DNA from a person selected at random contains the incriminating profile. No amount of effort and improved technology can reduce the error rate to zero, and the best protection a wrongly implicated, innocent person has is the opportunity for an independent retest. NRC II at 4 (emphasis added). See also id. at 24-25 (proposal to combine error rates and probability statistics is “ill-advised”). Because of this strong scientific opposition, courts that have considered the defendant’s argument have rejected it. See, e.g., Shea, 957 F. Supp. at 340, 344 (proposal for combining error rates with match probabilities is “seriously flawed because it would deprive the jury of the opportunity to determine the probability of a false match based on all of the evidence”); Copeland, 922 P.2d at 1320 (error rate not appropriately part of probability calculations); Kirkendahl, Order at 35-36 (Appendix I, Tab N)(rejecting Dr. Lawrence Mueller’s proposal to combine error rates and probability statistics because the NRC II called it “illadvised”); Dishmon, Order at xiii (Appendix I, Tab O)(same; adding that “no court decision from any jurisdiction in the United States adopting this method could be found”).73/ 73/ Even the few courts that have been sympathetic to the defendant’s other arguments have rejected the notion that lab error rates must be included in order to make probability calculations admissible. There is no need to calculate the lab error rate along 113 Thus, neither the NRC nor courts in other jurisdictions have required laberror rates to be included in probability calculations. Nor have any courts in the District of Columbia imposed such a requirement, although, in an abundance of caution, two of our trial courts have required that information about the results of proficiency tests be given during testimony about DNA test results. See Porter III, 1994 WL 742297, at *8; Bridgett, 120 Daily Wash. Law Rptr. 155, at 1704. Although the results of proficiency tests are not relevant to a Frye inquiry, we recognize that how well a lab performs generally accepted methods is a matter that the jury is entitled to consider. Porter II, 618 A.2d at 636. Accordingly, the FBI analyst who testifies at trial in this case will be prepared to address the forensic lab’s recent practices and experiences with proficiency tests. There is no need to address those issues before trial. Finally, the defendant also claims that, regardless of the scientific acceptance of the FBI's method for calculating probability statistics, the random match probabilities must be excluded because they are substantially more prejudicial than probative (Defendant’s Supplemental Motion at 27). That with the likelihood ratio from DNA analysis. Our system does not calculate lab error rates in any other forensic evidence effort. To include the two together disregards the empirical nature of scientific evidence. . . . Use of defense experts, discovery, and full cross-examination can address this form of error. Bokin, Trial Court Order at 12 n.9 (specifically rejecting Professor Koehler’s position that error rates be combined with match probability statistics). 114 argument was rejected by the Court of Appeals twelve years ago in Porter II. See 618 A.2d at 641 n.22 (refusing to hold that, merely because the numbers are so high, "admission of statistics as to the minuscule probability of a random match is unduly prejudicial"). To the extent that the PCR/STR analysis in these cases analyzed more loci than were typically examined in the RFLP process used in 1992, the counterbalancing probative value of the random match probability statistics has only increased since the Porter II decision. As Judge Richter has noted, “at its best, DNA is highly probative and accurate evidence.” Roberts/Veney Order at 18. Thus, the defendant cannot show that the probative value of the proffered statistics in this case will be substantially outweighed by the danger of unfair prejudice. See Johnson v. United States, 683 A.2d 1087, 1101 (D.C. 1996) (en banc). VII. The fact that the defendant was first identified by a “cold-hit” on a DNA database does not affect the admissibility of the DNA evidence. We demonstrated supra at 20-37 that the DNA evidence in this case was identified and analyzed by methods that are generally, if not universally, accepted in the relevant scientific community. We also showed that the FBI lab estimates the rarity of the DNA evidence profile, expressed as the Random Match Probability (RMP), based on statistical formulas that are also generally accepted in the relevant scientific community (see supra at 67-77). 115 The defendant nonetheless contends that this court must exclude the DNA evidence because he was first identified as a suspect by a search of a convicted offender DNA database, rather than by non-DNA evidence, and, so he claims, there is no generally accepted method by which to calculate the relevant probability statistics in such a “cold-hit” case.74/ He points to the existence of three separate approaches to presenting statistical information about cold hits as proof of a “raging controversy” among scientists that precludes general acceptance. The defendant is manufacturing controversy. His argument rests on three false premises: 1) there is only one question that statistics can answer in a coldhit case; 2) the existence of more than one approach to addressing a question means that none is generally accepted; and 3) a debate about the relative merits of different statistical approaches is the equivalent of a “raging controversy” over the fundamental reliability of those approaches. When defendant’s hyperbole is put aside, and the three statistical approaches he cites as evidence of controversy are examined more closely, it becomes clear that 1) there are at least two questions that statistics can answer in a cold-hit case; 2) 74/ Defendant raises this claim in a separate motion captioned “Motion To Exclude DNA ‘Inclusion’ Evidence, Expert Testimony, And Frequency Statistics Because There Is No General Acceptance As To The Statistical Methods Of Interpretation Of DNA Evidence That Derives From A ‘Cold Hit’ In A DNA Database.” Citations to that motion will be designated “Cold-Hit Motion.” 116 there is more than one generally accepted way of answering each question, although not all are equally favored; and 3) the issue defendant raises involves, at most, a debate over whether, in addition to being given a statistical estimate of the rarity of a DNA profile (Question #1), a jury also needs to be given a statistical estimate of the chance that the same DNA profile would be found in a database the size of the searched offender database (Question #2). Thus, the differences of opinion that defendant relies on are nothing like the controversy that arose in the early 1990s over whether the underlying assumptions about population genetics were sufficiently reliable to allow any calculation of probability statistics for a DNA match. See Conneally 9/04 Decl. at ¶15. None of the approaches defendant cites undermines the relevance of the estimated rarity of the DNA profile at issue (Question #1), or the previously demonstrated general acceptance and admissibility of the RMP to answer that question. See supra at 70-77. As for the admissibility of statistics concerning the probability of finding a match in an offender database (Question #2), the admissibility of numbers derived from the approaches defendant cites does not turn on their general acceptance in the relevant scientific community under Frye because all are accepted by scientists as valid ways to evaluate a match from a database search. Instead, their admissibility turns on the legal relevance of Question #2, and whether the proposed approaches to that question are compatible with legal evidentiary rule. Those determinations rest entirely within 117 the discretion of this court, and are not dependent on the approval of scientists.75/ 75/ The defendant tries to suggest that the court is required to apply a heightened admissibility standard to cold-hit cases “where the science itself is the lodestar of the government’s evidence against the defendant” (Cold Hit Motion at 3). That is not the law. Defendant relies on a passing comment by Judge Richter during the Roberts/Veney admissibility hearings. The Roberts and Veney cases did not involve cold hits, so Judge Richter’s comment is dictum at best. Moreover, the comment is directed toward cases in which the DNA match is the only evidence, which is not true in this case. See supra, at 3-6. Finally, the comment seems to conflate standards of admissibility with standards for deciding whether to allow a case to go to the jury or to grant a motion for judgment of acquittal. No court would refuse to admit otherwise admissible evidence, such as a fingerprint match, simply because there appeared to be no additional corroborating evidence and the court believed that, by itself, the fingerprint evidence was insufficient to support a conviction. No different standard should be applied to DNA evidence. 118 As we show, the first of the statistical approaches cited by defendant, which was recommended by the NRC I committee in 1992, is outdated and impractical, deprives the government of relevant evidence, and would mislead juries as to the estimated rarity of the DNA profiles. The two remaining approaches essentially reflect a disagreement over whether it is necessary to answer the question “what is the likelihood of finding the DNA profile from the evidence samples in a database of a certain size?” One school of thought, represented by the recommendation of the 1996 NRC II committee, assumes the legal relevance of that question and advocates providing a Database Match Probability (DMP) number to answer it. The other school, as described in an article by Professors Peter Donnelly and Richard Friedman, says that, although that question might be of interest in some contexts, and the NRC II formula is an acceptable way to answer it, that question is the wrong one to address in a criminal case.76/ This group believes, in effect, that no other probability figure needs to be introduced with, or instead of, the RMP. See Donnelly and Friedman, 97 Mich. L. Rev. at 962.77/ 76/ See Peter Donnelly and Richard Friedman, “DNA Database Searches and the Legal Consumption of Scientific Evidence,” 97 Mich. L. Rev. 931 (1999), attached in Defendant’s Appendix VII, TAB J. Other proponents of this view include David Balding, A.P. Dawid, Ian Evett, J. Mortera, and Bruce Weir. 77/ Donnelly and Friedman also point out that a database search produces additional information that a fact finder might find probative; that is, one DNA profile in a database was found to 119 match DNA from the evidence while many other profiles were found not to match. Donnelly and Friedman, 97 Mich. L. Rev. at 962. To illustrate, in this case, even though the Virginia database search was conducted on only an 8-loci profile, only one match with the DNA from the evidence samples was found and that match was with the defendant’s DNA. That means that over 100,000 former offenders’ DNA did not match the DNA from the blood at the crime scene and, therefore, over 100,000 people have been eliminated as the possible source of that blood. Even defendant’s experts agree that this information has probative value. Both Doctors Mueller and Krane testified on this issue at a DNA admissibility hearing in State v. Robinson, Case No. 00F06871 (Superior Court, County of Sacramento, State of California). See excerpts of their testimony, Appendix VII, TAB H, at pp. 1571 (Mueller). 120 The government intends to introduce only the RMP numbers that represent the estimated frequency with which the 13-loci DNA profile found in the evidence samples would be expected to occur in four major population groups. See supra at 6-9.78/ We believe that the rarity of the DNA profile found in the evidence samples is always relevant. Presentation of the RMP alone does not risk suggesting potentially prejudicial information about the defendant’s criminal record. In addition, presentation of the RMP is consistent with the practice of almost every DNA forensic laboratory in the country. See Budowle Decl. at ¶12; Sensabaugh Decl. at ¶ 21 79/ 78/ The government has no intention of relying on the match between the DNA from the evidence samples and the defendant’s 8loci DNA profile in the Virginia offender database, unless the defendant opens the door to admission of that evidence. All of the DNA evidence that the government will introduce in its case-inchief is based on the subsequent confirmatory re-testing and examination by the FBI laboratory, which found a match between the DNA from the evidence samples and defendant’s DNA at 13 loci. 79/ In support of our opposition to defendant’s cold-hit argument, we have included in Appendix VII, the declarations of four experts: Dr. Bruce Budowle, the Senior Scientist in the Laboratory Division of the FBI, and one of the pioneers in the development of DNA analysis for forensic identification purposes (TAB A); Dr. Michael Conneally, Distinguished Professor of Medical Genetics and Neurology at Indiana University Medical Center, whose specialization is the statistical aspects of population genetics (TAB B); Dr. James Crow, Professor Emeritus of Genetics at the University of Wisconsin, who chaired the National Academy of Sciences/National Research Council’s Committee on DNA Technology in Forensic Science, which published the 1996 report “The Evaluation of Forensic DNA Evidence” (NRC II) (TAB C); and Dr. George Sensabaugh, Professor of Forensic Biology in the School of Public Health at the University of California at Berkeley, who was the 121 However, if the court believes that the likelihood of finding a match to the evidence profile in the offender database is relevant, we will not object if the defense wants to introduce, in addition to the RMP figures, the DMP figure as calculated by the generally accepted formula recommended by the NRC II committee in 1996. There is no controversy in the relevant scientific community that the use of the NRC II formula is both a correct and conservative (as well as a defendant-favoring) way to answer that question. The government objects to admission of statistics based on the NRC I approach because it would unjustifiably keep from the jury the fact that the defendant’s DNA matches the evidence profile at the 8-loci used in the Virginia database search, in addition to matching the five remaining loci on the full 13loci matching profile. only scientist to serve on both of the special committees on the forensic use of DNA convened by the National Research Council of the National Academy of Sciences (commonly known as NRC I and NRC II) (TAB D). Because we have also included in our appendix, in support of other arguments, the declaration that Dr. Conneally submitted in January 2003 in United States v. Roberts and United States v. Veney, we will identify his declaration concerning cold-hit issues as “Conneally 9/04 Declaration”. 122 A. The Random Match Probability (RMP) is always relevant and admissible to show the rarity of a DNA profile. The RMP figure that the government intends to introduce with the DNA evidence measures the estimated frequency, or rarity, with which a particular DNA profile is expected to occur in a given population. The rarity of a profile is relevant information, and can be highly probative. If the person who commits a crime is believed to have a particular characteristic - whether it is height, eye color, blood type, or a DNA pattern – then it is relevant that the defendant also has that characteristic, because sharing that characteristic tends to make it more probable that the defendant is the person who committed the crime. See Winfield v. United States, 676 A.2d 1, 2 (D.C. 1996) (en banc) (evidence is relevant if it tends “to make the existence or nonexistence of a fact more or less probable than would be the case without that evidence”) (internal quotation marks and citation omitted). If the characteristic is fairly common, then the fact that the defendant also has that characteristic, while still probative of identity, will nonetheless carry relatively little weight. Conversely, if the characteristic is very rare, then the fact that the defendant has that characteristic can be highly probative. The relevance of the RMP as an indicator of the rarity of the DNA profile found in an evidence sample is the same regardless of whether the suspect is identified by non-DNA evidence or by a DNA database search. Crow Decl. at 123 ¶9. In fact, the RMP can be calculated from the DNA profile in the evidence sample before any suspect is identified. Sensabaugh Decl. at ¶ 21. Once determined, the RMP of a DNA profile never changes, unless additional loci are examined. Crow Decl. at ¶11; Sensabaugh Decl. at ¶ 21.80/ It is the common practice of forensic laboratories to report the RMP regardless of whether the defendant was first identified from a database search or from non-DNA evidence. See Budowle Decl. at ¶12; Sensabaugh Decl. at ¶ 21. Although not all cases involve a 13-loci match,81/ in a case like this one where the DNA profiles from the evidence samples and from the defendant match at 13 loci, the probative value of the match is very high. A profile containing a particular combination of 26 alleles (13 loci X 2 alleles at each locus) is an exceedingly rare event in statistical terms. See Crow Decl. at ¶¶ 13, 16. It is so rare that the government’s expert is prepared to opine that, to a reasonable degree of scientific certainty, the DNA profile is unique. See supra at 8. The jury is entitled to learn of the rarity of the full evidence profile, in order to give proper weight to the significance of a match at all 13 loci. The value of the RMP should not be artificially diminished. 80/ For an explanation of the generally accepted formulas used to develop the RMP in cases involving single-source DNA samples and mixed-source DNA samples, see supra at 77-91. 81/ Sometimes there is insufficient DNA from which to obtain a 13loci profile. 124 The three statistical approaches described in defendant’s motion are not substitutes for the RMP. Those approaches focus on an entirely different question, and the way in which one chooses to address that question does not affect the relevance and general acceptance of the RMP as a measure of the rarity of a DNA profile. The DNA Advisory Board (DAB) explained in a 2000 position paper that there often are several statistical approaches to evaluating DNA evidence, and all may be valid, depending on the question being asked: When a comparison of DNA profiles derived from evidence and reference samples fails to exclude an individual(s) as a contributor(s) of the evidence sample, statistical assessment and/or probabilistic reasoning are used to evaluate the significance of the association. Proper statistical inference requires careful formulation of the question to be answered, including, in this instance, the requirements of the legal system. . . . As the NRC II Report (1996) describes,82/ there are alternative methods for assessing the probative value of DNA evidence. Rarely is there only one statistical approach to interpret and explain the evidence. DNA Advisory Board, “Statistical and Population Genetics Issues Affecting the Evaluation of the Frequency of Occurrence of DNA Profiles Calculated From Pertinent Population Database(s)” (February 23, 2000), published in Forensic Science Communications Vol. 2 No. 3 (July 2000) (hereafter “DAB Position 82/ See NRC II at 192. 125 Paper”) (Appendix VII, Tab E).83/ After repeating its general endorsement of the NRC II report, the DAB explained that its position paper was written to clarify issues that arose in special cases, including cases involving a felon database search. DAB Position Paper at 2. With respect to such cold-hit cases, the DAB said: Two questions arise when a match is derived from a database search: (1) What is the rarity of the DNA profile? and (2) What is the probability of finding such a DNA profile in the database searched? These two questions address different issues. That the different questions produce different answers should be obvious. Id. at 6. The first question, which is relevant in all cases involving DNA evidence, “addresses the random match probability, which is often of particular interest to the fact finder.” Id. See Budowle Decl. at ¶11; Conneally 9/04 Decl. at ¶7; Crow Decl. at ¶9; Sensabaugh Decl. at ¶¶ 21-22 . The second question – what is the probability of finding such a DNA profile in the database searched? – is the question at issue in the three approaches cited by defendant. B. The approaches cited by defendant are generally accepted methods of answering questions other than the rarity of a DNA profile, but not all are admissible under legal evidentiary standards. VII. The NRC I Approach 83/ Defendant includes an incomplete excerpt from this position paper in the Appendix to his Cold-Hit Motion. The government has included the entire paper so that its meaning is not misunderstood. 126 The first, now discredited, approach to evaluating the relevance of a cold hit on a database was recommended by the NRC I committee on forensic DNA evidence in 1992. It is generally accepted in the relevant scientific community that the NRC I approach is outdated and unnecessarily wastes probative information. For the same reasons, use of the NRC I formula is not legally appropriate because it leads to the unjustifiable exclusion of probative information. The NRC I report was issued more than a dozen years ago when DNA testing methods were less reliable and less discriminating, and therefore scholars feared that they would be more likely to lead to purely coincidental matches with innocent people. See Conneally 9/04 Decl. at ¶12; Crow Decl. at ¶12. The NRC I committee reacted by recommending an approach that keeps from the jury not just the statistic based on the database match, but the fact of a database match and any evidence that was used to obtain the database match. NRC I recommended testing a limited set of DNA markers for comparison with a database. If there was a match, then the evidence sample and the DNA of the person who matched would be re-tested on an entirely different set of markers. Only the second set of markers would be used to develop the random match probability number, and the jury would not be told that the defendant’s DNA profile also matched the evidence profile at the first set of markers. See NRC I at 124. The NRC I approach was developed before there had been any cold hits on DNA databases, and at a time when DNA testing was conducted by the restriction fragment length polymorphism (RFLP) method. Sensabaugh Decl. at ¶ 17. RFLP tended to examine a much smaller number of DNA markers than the modern system of PCR/STR analysis, and was less precise, which meant that the possibility of a purely coincidental match was greater. The NRC I 127 committee recognized that there were alternative statistical approaches to evaluating the effect of database searches; however, it rejected those approaches because of doubts about the ability under then-current testing methods and sampling techniques to accurately estimate genotype frequencies. NRC I at 124. Instead, it chose to avoid the database issue, by ignoring the markers used in the initial database search. NRC I at 124. A mere four years after the NRC I report was published, and before forensic laboratories had applied the NRC I recommendation, the NRC II committee concluded that the NRC I approach wasted information, and would create problems in cases in which there was insufficient DNA for subsequent re-testing. NRC II at 34, 133. It recommended instead the reporting of a database match probability (DMP) number, which is described infra at 112-116. In addition to the deficiencies noted by the NRC II committee, the NRC I approach would be costly and impractical. To reduce the possibility of an erroneous identification of a suspect from a database, it would be desirable to use a significant number of validated markers in a database search. The NRC I approach would require laboratories to develop and validate a similarly large set of independent markers for use at trial. Maintaining the ability to test and analyze DNA at numerous markers, and identifying and validating new markers for such testing, would be very time-consuming and costly. See Donnelly and Friedman, 97 Mich. L. Rev. at 965 & n.85 (describing additional testing and validation costs). Moreover, the NRC I approach is no longer appropriate. Even if, from a statistician’s perspective, that approach may still be a technically acceptable way of addressing a database search, it is generally agreed among knowledgeable scientists that the current ability to develop more accurate and discriminating DNA profiles makes the NRC I recommendation outdated and 128 unnecessarily conservative. See Conneally 9/04 Decl. at ¶12; Crow Decl. at ¶15; Sensabaugh Decl. at ¶18. Thus, like the ceiling principles that were adopted by the NRC I committee to compensate for uncertainties about the reliability of the population genetics on which the statistical probabilities were based, and that were later abandoned as further research and experience erased those uncertainties, see supra at 70-77, the NRC I cold-hit case recommendation has outlived its arguable relevance. Defendant’s assertion that the NRC I approach “currently enjoys a number of followers” does not withstand scrutiny. Cold-Hit Motion at 18. He cites the writings of Professors Newton Morton, Richard Lempert, and Aidan Sudbury from seven or more years ago, before forensic laboratories were widely using the more discriminating PCR/STR methods of developing and analyzing DNA profiles. The fact that scientists might have agreed with the NRC I recommendation years ago is not proof that they continue to hold those views today, particularly in light of the shortcomings that were identified by the NRC II report, and the significant advances in the science of DNA analysis since both NRC reports were published. For example, Professor Richard Lempert, a member of the NRC I committee who often serves as a criminal defense attorney, and who once agreed with the NRC I recommendation, now rejects that approach. In Professor Lempert’s opinion, “the recommendation of NRC I is outdated and defunct [, and] there is no need to type additional loci when a cold hit has been made from a felon database. This is based on the fact that the current STR technology provides sufficient power of discrimination.” Declaration of Dr. Richard Lempert [in People v. Robinson, 129 No. 00F06871, Superior Court for the County of Sacramento]; see App. VII at Tab G.84/ For similar reasons, Dr. George Sensabaugh, the only scientist to serve on both NRC I and NRC II, also rejects the NRC I recommendation. Sensabaugh Declaration at ¶18. Moreover, the fact that some scientists may prefer this method does not mean that they think that other methods are not generally accepted. For example, a scientist might well say that testing blood samples is a more accurate and better way of assessing the degree of alcohol intoxication of a suspect. Even if true, it would not follow that use of a Breathalyzer test would not be generally accepted as a sufficiently reliably way to assess the same thing. In the same vein, although Dr. Newton Morton has written in support of the NRC I approach, he also agrees in principle with the NRC II approach; he does not endorse that approach because of concerns that it would reveal to the jury prejudicial information about a defendant’s criminal record. See Donnelly and Friedman, 97 Mich. L. Rev. at 963 n.81.85/ 84/ It is surprising that defendant claims that Professor Lempert still endorses the NRC I position (see Cold Hit Motion at 18). As indicated by Professor Lempert’s declaration, he found it necessary to correct that misrepresentation of his views in a California case captioned People v. Robinson. Professor Lempert was responding directly to inaccurate assertions about his position that apparently had been made in the Robinson case by two of the defendant’s proffered experts in this case – Dan Krane and Laurence Mueller. Neither one repeats those inaccurate assertions here. 85/ It is worth noting that this reason for preferring NRC I has nothing to do with the validity of the science, but relates to legal concerns about the prejudicial effect of information on the jury. 130 Nor is it significant that defendant’s experts – Dan Krane and Laurence Mueller – prefer the NRC I approach.86/ Those defense experts advocated the same position in a recent California case that dealt solely with the cold-hit issue. Their position was rejected because the trial court concluded that they were not prominent experts in the relevant field, and their views were not representative of such experts. After considering the testimony of five expert witnesses, a Sacramento, California, trial court found that it was generally accepted in the relevant scientific community to present just the RMP as a measure of the significance of a DNA match in a cold-hit case. See Transcript of Court’s Ruling Re: DNA Evidence, People v. Robinson, No. F06871 (Superior Court of the State of California, County of Sacramento, Dept. 30, Feb. 20, 2003) [Honorable Peter Mering presiding] (hereafter “Robinson Ruling”) at 21 (included in Govt App. VII, Tab F).87/ The defendant in Robinson relied on the testimony of Drs. Krane and Mueller, who, as they have in this case, attempted to persuade the court that there was a bitter controversy, and therefore no consensus, about the correct way in which to present a statistic evaluating the significance of a 86/ The defendant also presents the declaration of Dr. Sandy Zabell, but Dr. Zabell simply describes the relationship among the three approaches described in the text. He does not suggest that any of three approaches is wrong or not generally accepted, or that the existence of three approaches is evidence of a lack of consensus as to their general acceptance in the scientific community. 87/ Although California adds requirements to the Frye test for the admissibility of evidence derived from novel scientific methods, the first step in its analysis is the general-acceptance standard announced in Frye, which still applies in the District of Columbia. See supra at 16-20. 131 match in a cold hit case. They also advocated use of the NRC I approach. The trial court rejected their opinions, finding that: [W]hile [Drs. Krane and Mueller] are qualified, . . . they are [not] heavy hitters in the scholarly world. They are not persons whose articles are quoted or considered authoritative in the field. And it is a fact of some importance that they have consistently over the years resisted virtually each and every development in DNA processing and DNA science, and continue to resist many of them even though [the developments] have, as time has passed, become accepted by the general scientific community. So . . . either they are of the old school in some way are extremely conservative in the way these things ought to appear, or they have discovered that their services are in demand because they do adhere to the old system and reject the new. So I don’t feel their position is one of substantial scholarly importance. Robinson Ruling at 14.88/ 88/ The California court’s conclusion was amply supported by the record. For example, Dr. Mueller admitted that he does not attend professional conferences at which current developments in DNA analysis and interpretation are presented and discussed, does not publish on forensic DNA subjects, and has testified for the government only once in all of the 125-160 cases in which he has previously testified (see App. VII, TAB H a pp. 1527, 1537-38). Similarly, Dr. Krane does not attend relevant professional conferences, is not a participant in working groups on DNA analytic methods, has never published on the cold hit issue, and testifies almost exclusively for defendants (App. VII, TAB H, at 1770-73, 1777-78). He advocates positions, such as use of the ceiling principle, that he acknowledges are not generally accepted in the relevant scientific community. App. VII, TAB H at 1794-1795. In another example, although Dr. Krane agrees that the rarity of the tested DNA profile is relevant, and that the NRC I formula does not give the “closest true frequency” of such a profile, he still endorses the NRC I approach. Id. at 1873-1874. Dr. Krane candidly admitted his bias toward proposals that favored the defendant in the Robinson case, agreeing that it was his “inclination in every criminal case, . . . that we should provide conservative statistics that benefit the defendant.” App. VII, TAB H at 1795. 132 Finally, even if there were some scientists who still preferred the NRC I approach, that is not a reason to choose that approach over other approaches that reflect more up-to-date science, and do not result in the exclusion of probative evidence. As the Court of Appeals emphasized in Porter II, “[I]f the evidence offered conduces in any reasonable degree to establish the probability or improbability of the fact in controversy, it should go to the jury.” United States v. Porter, 618 A.2d 629, 640 (D.C. 1992). Thus, the NRC I approach unjustifiably would hide from the jurors in this case the fact that the defendant’s DNA profile matched the DNA from the evidence samples at the eight markers initially used to make the database match. It would also prevent the jurors from learning of the RMP based on the entire 13-loci profile, and leave them with an artificially reduced RMP that would give them the false impression that the DNA profile is more common than it probably is. Applying the NRC I approach to non-DNA evidence illustrates the absurdity of its result in a criminal case. Suppose that a woman is robbed by a person whom she is able to describe to the police in detail as: 1) male; 2) fair complected; 3) bald, but with 4) grey hair around the sides; 5) blue-eyed; 6) with a moustache; 7) with ears that stick out; and 8) over 40 years old. The victim also tells the police that the robber 9) held his gun in his left hand; 10) had a deep voice; 11) was bow-legged; 12) stood about 5'11" tall; and 13) weighed about 300 lbs. After trawling through a book of mug shots, the victim selects a photograph of a man who looks like the robber based on the first eight characteristics that she described. To confirm that identification, the police have her look at a line-up which includes that man. The line-up is arranged so that the victim can only see the individuals in silhouette; that is, she is deliberately prevented from seeing any of the 133 characteristics by which she selected the photograph. The silhouettes reveal height and weight, and, through demonstrations, the witness is able to observe whether the individuals in the line-up are bow-legged, are left-handed, or have deep voices. Based on those five characteristics, the victim selects the same man that she identified from the photographs. At trial, the government is allowed to present evidence about the line-up identification, but it is precluded from mentioning the photo identification or that the victim initially described 13 rather than five characteristics of the robber, and the victim is permitted to testify only that the robber was 5'11" tall, about 300 lbs, bow-legged, and left-handed, and had a deep voice. Obviously such a distorted procedure could not be justified with respect to non-DNA evidence. Nor can it be justified for DNA evidence on the basis of an outdated recommendation. Defendant’s suggestion that the NRC I approach is justified because there is an increased danger that an innocent person will be identified in a database search is misguided. The discriminating power of a DNA profile increases as the number of markers included in the profile increases. There is a greater likelihood of finding several persons whose DNA profiles match if only four or five markers are compared (as was typically the case when the NRC I and NRC II reports were issued). Sensabaugh Decl. at ¶ 17. That possibility decreases as additional markers are compared until the likelihood of a coincidental match is infinitesimally small.89/ The DNA 89/ Thus it is not surprising to learn of a 9-loci match, and certainly not surprising to learn of several 6-loci matches, particularly where most of the 6-loci matches were among brothers. Siblings are more likely to share similar DNA profiles, but even they are not likely to be found to match at a higher number of loci. Nothing suggests that relatives of the defendant should be suspected of the murder in this case. 134 profiles developed from the numerous evidence samples in this case were, with few exceptions, 13-loci samples. See supra at 6-9. To date, genetic researchers have never seen a false coincidental match between 13-loci profiles. Crow Decl. at ¶13. Defendant correctly notes that the NRC I approach “could dramatically reduce the government’s random match probability calculations by a number of magnitudes.” Cold-Hit Motion at 17. This, in turn, would give the jury the false impression that a given DNA profile was “dramatically” more common than it likely would be if all the matching markers were used to calculate the RMP. There is no scientific imperative that requires the court to admit an artificially deflated RMP, or keep from the jury the fact that the defendant’s DNA matched the DNA from the evidence at the eight loci used in the database search. The NRC I approach unjustifiably deprives the jury of otherwise probative and relevant evidence. Even if the NRC I approach was not practically obsolete, statistics derived from the NRC I approach should not be admitted. 2. The NRC II Approach A second approach was recommended by the NRC II committee in 1996. That committee noted that NRC I’s approach wasted information unnecessarily and assumed that DNA samples could be retested, which was not always the case. NRC II at 124. In its Recommendation 5.1, NRC II recommended a method that uses information about all of the tested markers to develop the RMP, then multiplies that RMP by the number of profiles searched in the database, or “N”. The result is a number that represents the chance of finding a match – whether true or coincidental – in a database of a certain size. That number is sometimes called the Database Match Probability (DMP). 135 Knowledge of the DMP does not help the fact finder decide whether the match is true or coincidental. However, if you assume that the true source of the DNA evidence sample is not in the database, then the DMP can answer the question “what is the chance that the match between the DNA evidence profile and the profile found in the database is coincidental?” Expressed that way, the DMP might be information that the defense would want to introduce to a jury. The DMP does not measure the rarity of a profile, however, and thus it is not a substitute for the RMP in answering that question. The rarity of a profile is typically the more relevant question. For that reason, forensic laboratories always report an RMP number in cold hit cases, and provide a DMP number if it is requested by one of the parties. See Budowle Decl. at ¶12.90/ Although some commentators have treated the NRC II recommendation as though it precludes giving the RMP, the NRC II committee did not say that, if the DMP was given, the RMP should not also be used to express the rarity of the DNA profile. Even if the language of Recommendation 5.1 is ambiguous in some respects, the subsequent DAB position paper clarifies that the NRC II formula answers only one of two questions that arise in a cold-hit case. The RMP answers the other.91/ 90/ Defendant is misinformed when he insists that “no scientist agrees with the government’s proposal to simply place before the jury the random match probability [in a cold hit case] as one does in more traditional DNA ‘confirmation’ cases.” Cold Hit Motion at 11. Forensic scientists throughout the country do just that. Moreover, that is essentially the effect of following the third approach cited by defendant. See infra at 116-118. 91/ Defendant mischaracterizes by omission one of the important points made in the DAB position paper, which is that there are two relevant questions in a cold hit case. Although the paper expressly concerned itself with two of the alternative proposals for addressing the likelihood of finding a DNA profile in a 136 database of a particular size, that does not mean that the DAB thought that the rarity of a profile was unimportant. To the contrary, the DAB described that question as “often of particular interest to the fact finder.” DAB Position Paper at 8. 137 Dr. George Sensabaugh, who is the only individual to have served on both NRC I and NRC II, explains that, based on the language of the report and presentations and discussions at NRC II meetings, it was not his understanding that the NRC II recommendation to report a DMP was intended to preclude introduction of the RMP to express rarity of the profile. Sensabaugh Decl. at ¶22. Nor was that the understanding of Dr. James Crow, a prominent population geneticist who chaired the NRC II committee. Crow Decl. at ¶11. That is the only logical way to read the NRC II report as a whole because, as even defendant’s experts agree, the DMP and RMP measure different things, see App. VII, TAB H, pp. 1568 (Mueller), and only the RMP measures rarity. See Conneally 9/04 Decl. at ¶7. In addition, the fact that a suspect has been identified through a database search does not mean that the DNA profile of interest has become more commonplace. All that a database search does is make it more likely that you will find the profile that you are looking for, but that would be true regardless of whether the profile was common or rare. Moreover, providing just the DMP number to a jury would be misleading. For example, if the RMP for an evidence profile was 1 in 1,000,000, that tells the jury that the estimated frequency with which the profile would be expected to occur in a particular population is one person in a million. See Budowle Decl. at ¶15. Suppose we then searched for that profile in a database of 1 Million people. The DMP would be calculated by multiplying the RMP of 1/1,000,000 by the size of the database (1,000,000). The resulting DMP would be 1, which suggests that you would expect to find that DNA profile in that database one hundred percent of the time. If that were the only figure given to the jury, it could easily be misinterpreted as meaning that everyone in the population has that DNA profile, or at least that the profile was 138 extremely common, when in fact, at least 999,999 people were actually found not to have that profile. Budowle Decl. at ¶15. Such a distortion of the evidence is unwarranted, and can be avoided by ensuring that, in any case in which the DMP is introduced, the RMP is also provided as a separate measure of the rarity of the profile. That is essentially what would occur in a cold hit case if the position of the DNA Advisory Board was followed. In its 2000 position paper, the DAB examined the arguments for and against presenting a statistic that expresses the chance of finding the DNA profile of interest in an offender database. It chose to endorse the NRC II recommendation that a DMP number be given, primarily because that was a conservative approach. DAB Position Paper at 8; see also Budowle Decl. at ¶18.92/ At the same time, the DAB would not oppose giving the RMP as well, because the DMP measures only the probability of finding a profile in the database, and does not supersede the need to introduce the RMP as a measure of the profile’s rarity. See Budowle Decl. at ¶10.93/ 92/ Dr. Budowle was one of the authors of the DAB Position Paper. Budowle Decl. at ¶10. 93/ The DAB specifically considered whether the NRC II recommendation or the third approach cited by the defendant (the Donnelly and Friedman approach) was preferable in a cold hit case. The Board ultimately chose to endorse the NRC II approach because it was more conservative. In its position paper, the DAB points out that the reasoning of the Donnelly & Friedman approach depends heavily on an understanding of the use of likelihood ratios and Bayesian logic, concepts that are difficult for laymen to understand and apply in a criminal setting. Thus, the DAB position paper concludes that, even though the NRC II approach and the Donnelly/Friedman approach both “yield a coherent evaluation of the evidence” in the context of likelihood ratios, the Board prefers the NRC II approach because it “communicates value of a database search far better, and it is always conservative.” DAB 139 3. The Donnelly and Friedman Approach Adherents to the third approach identified by the defendant agree with the NRC II committee that the NRC I recommendation is unnecessarily wasteful and impractical. See, e.g., Donnelly & Friedman, 97 Mich. L. Rev. at 963-966. However, they also argue that it is unnecessary to treat a cold hit case differently from a case in which the defendant is first identified by non-DNA evidence. Position Paper at 8; see also Budowle Decl. at ¶18. 140 The third approach (hereafter “the Donnelly & Friedman approach”) reflects the intuitively logical idea that “the value of a DNA match is attributable to the rarity of the profile.” Donnelly and Friedman, 97 Mich. L. Rev. at 944.94/ The question of interest to the jury is how probable is it that the defendant whose profile matched the evidence sample profile is in fact the source of that sample, and that question can be answered in a simple way by the RMP. Id. at 947.95/ It is undisputed that the fact that the defendant is first identified through a database search does not change the rarity of the DNA profile at issue, as expressed by the RMP. Thus it does not change the RMP’s value in proving that the defendant is the source of the DNA found in the evidence. Donnelly & Friedman also point out that, if anything, the fact that the defendant’s profile was the only profile found to match the evidence while many other profiles were found not to match makes the evidence overall slightly more probative than the RMP alone. Id. at 948.96/ 94/ Other notable scientists who generally agree with the position described by Donnelly and Friedman include David Balding, A.P. Dawid, Ian Evett, J. Mortera, and Bruce Weir. Conneally 9/04 Decl. at ¶14. 95/ Although the proponents of the Donnelly & Friedman approach typically rely on more complex ways of presenting statistical information than simply statements of the RMP (such as likelihood ratios and Bayesian logic), the logic of their reasoning is that the RMP expresses the information that the jury needs to hear. See Donnelly & Friedman, 97 Mich. L. Rev. 96/ Contrary to the defendant’s repeated assertions, Donnelly & Friedman do not contend that, under their approach, the overall value of the DNA evidence is significantly greater in a cold hit case than in other cases. Therefore, their approach is far more moderate, and the differences with the NRC II approach are far less, than the defendant would have this court believe. 141 To the extent that people interpret NRC II as dictating only the DMP, Donnelly and Friedman also underscore the anomaly of the NRC II approach as the size of the searched database approaches the size of the entire population in question. Basically, under the NRC II formula, the probative value of the DNA match appears to decrease as the size of the database, or the number of profiles actually searched, increases. Eventually, if all profiles in the population are searched, and only one profile matches the evidence profile, then the logic of the NRC II formula leads to the conclusion that the unique match has no probative value. Of course, if everyone else in the population has been eliminated as a source of the evidence profile, then the match should instead be conclusive proof of identity. See Donnelly and Friedman, 97 Mich. L. Rev. at 952-954. The NRC II report recognizes this anomalous result, but offers no solution. See NRC II at 161. The Donnelly and Friedman article presents cogent criticisms of the NRC II’s view that a database match probability number must be presented to a jury, but this does not create a controversy under Frye. Donnelly and Friedman emphasize that they are not saying that the NRC II recommendation is based on unacceptable science. Rather, they contend that NRC II answers an unnecessary question in a forensic setting. The question tested by the NRC II approach is whether the source of the crime sample is in the database, a question that may intrigue academicians, but is of little or no interest to a jury. Id. at 946. At bottom, Donnelly & Friedman are arguing legal relevance, not scientific reliability. At least one trial court to have addressed the various statistical approaches in a cold hit case has found, under a Frye-Plus test, that the Donnelly and Friedman approach is generally 142 accepted in the relevant scientific community. See Robinson Opinion, App. VII, TAB F.97/ The result is that the government will be allowed to introduce the RMP. Id. At the same time, that court also held that the defendant would be permitted to introduce the DMP figure, as calculated by the NRC II formula, if he chose to do so. Id. Because whether the DMP should be introduced in addition to the RMP is a matter of legal relevance rather than scientific acceptance, that decision is an acceptable compromise between the NRC II approach and the Donnelly and Friedman approach. Therefore, as previously noted, if this court believes that a measure of the chance of finding the evidence profile in a database is relevant, and the RMP is admitted as a measure of the rarity of the DNA profile, then the government has no objection to the defendant introducing the DMP as calculated by the generally accepted and conservative NRC II formula.98/ Where the RMP is also given, the DMP cannot be misinterpreted, and the introduction of both would be a scientifically and legally acceptable, albeit conservative, approach to explaining the significance of a database match. Where there is a choice between two differing approaches, but agreement that one is acceptably conservative, then the court does not err by following the more conservative path. Cf. Porter, 618 A.2d at 640-642. 97/ Applying a Frye-Plus test, a Minnesota trial court has also recently rejected, without a hearing, the argument that a different statistical rule than the RMP should be applied when a suspect is identified from a database search. See State v. Johnson, Court File No. 02096915, and State v. Grant, Court File No. 04031951, District Court, Hennepin County (August 16, 2004) [Honorable Steven A. Pihlaja]; attached at Appendix VII, Tab I. 98/ This could be done through cross-examination of government’s expert or through a separate defense expert. 143 the Of course, if the DMP figure is admitted in addition to the RMP, the court should bear in mind the likely need to sanitize references to the database (as is often done, for example, when arrest photos are introduced as part of a photo array); and the fact that introduction of the database figure will open the door to appropriately sanitized cross-examination (or redirect) about how the DMP was calculated, and how many individuals were actually excluded as sources of the DNA from the crime scene evidence as a result of the database search. C. The failure to incorporate estimates of error rates into the probability statistics does not preclude admission of the DNA evidence in a cold hit case. As he did in his Supplemental Motion with respect to all probability figures, defendant contends that the probability figures used in a cold-hit case must incorporate estimates of laboratory error, or “false positive,” rates before the probability statistics are admissible (Cold Hit Motion at 2426). We have previously demonstrated that this issue is not properly before the court because, ordinarily, the question of error goes to the weight of the evidence and not its admissibility. See supra at 91-94. In addition, such an approach is generally rejected by scientists. Id. For the same reasons, defendant’s argument does not preclude the admissibility of the DNA evidence.99/ 99/ Ironically, defendant cites in support of his error-rate argument Professor David Balding’s 1997 article, which was written in reaction to the NRC II report (Cold Hit Motion at 24). In that article, Professor Balding agrees with the government’s position that it is inappropriate to combine error rates with match probabilities. See Balding, “Errors and Misunderstandings in the Second NRC Report,” 37 Jurimetrics J. 469, 476 (1997) (attached in defendant’s Cold Hit Appendix). He also concludes that “what matters are not the probabilities of any profiling or handling errors, but only the probabilities of errors that could have led to the observed DNA profile match.” Id. at 475-476. 144 In addition, defendant misses the mark when he suggests that the court should be concerned about the possible error rate at the labs that administer the Virginia database program (Cold Hit Motion at 26). The evidence that the government intends to introduce is derived solely from the confirmatory testing by the FBI lab, which involved re-testing DNA from a new sample of defendant’s blood. The practices of the Virginia lab are irrelevant. D. The probative value of the DNA evidence in a cold-hit case is not outweighed by prejudicial effect. Similarly, defendant repeats the argument that the probative value of the DNA evidence is outweighed by its prejudicial effect, but his arguments are unpersuasive (see Cold Hit Motion at 2628). As we noted earlier in our opposition, the Court of Appeals in Porter rejected the argument that jurors will be so overwhelmed by statistics suggesting “a minuscule probability of a random match” that they cannot be trusted to consider the evidence. Porter II, 618 A.2d at 641 n.22. Except for adding arguments about how error rates should reduce the probative value of random match numbers, defendant’s Cold Hit Motion adds nothing else new to his argument about prejudice. That argument should be rejected for the reasons we have stated before. See supra at 90-94. Defendant also overstates the conclusions of an article by Professor Jonathan Koehler, by suggesting that it supports the idea that “the power of the DNA evidence to mislead and confuse the jury is great” (Cold Hit Motion at 27). The primary thrust of the article is how jurors’ perceptions of DNA evidence can be changed by the way in which the prosecutor and defense attorney describe it, which, to a trial attorney, makes the DNA evidence no different from other kinds of evidence. 145 E. The chain of custody for defendant’s DNA sample in the Virginia offender database is irrelevant to this case. Defendant’s final complaint is that the United States cannot prove the chain of custody for the sample of biological evidence from defendant that was entered into the Virginia offender database (Cold Hit Motion at 28-29). The chain of custody in Virginia is irrelevant because, after the cold hit on the Virginia database, the United States obtained a new sample from appellant and the FBI performed a confirmatory test on that new sample. The defendant does not challenge the chain of custody for that confirmatory sample. Even if defendant had challenged the chain of custody for the relevant DNA samples, there is no need to consider chain of custody issues before trial. The cases defendant cites stand for the unremarkable proposition that, where the government intends to rely on real evidence that is fungible, it must be able to establish, through a chain of custody, that the evidence offered at trial is connected to the defendant. E.g., Turney v. United States, 626 A.2d 872, 843 (D.C. 1993) (government required to prove that material recovered from defendant was same material analyzed by chemist and found to be illegal drug); Novak v. District of Columbia, 82 U.S. App. D.C. 95, 95-96, 160 F.2d 588, 588-89 (D.C. Cir. 1947) (same regarding urine sample and urinalysis test results); Smith v. Deppish, 807 P.2d 144, 238 (Kan. 1991) (noting, in dictum, that DNA test results that are otherwise generally accepted in scientific community may still be challenged on evidentiary grounds, such as those involving chain-of-custody); State v. Scott, 33 S.W.3d 746, 760 (Tenn. 2000) (“witness must be able to identify the evidence or establish an unbroken chain of custody”; also noting that chain of custody was 146 “especially important” in cases, like Scott’s, involving mitochondrial DNA (“mtDNA”) because mtDNA is “hypersensitive” to contamination).100/ Ordinarily, “evidence of a break in the chain of custody only affects the weight to be given to the evidence.” In re D.S., 747 A.2d 1182, 1187 (D.C. 2000). “[W]hen physical evidence is in the hands of government, the presumption arises that it has been handled properly.” Id. Moreover, the government does not have to prove such a chain beyond any doubt; “once the government has established an ‘unbroken chain of custody as a matter of reasonable probability,’ defendant must present evidence of tampering” before the challenged evidence may be excluded. Id. (internal citation omitted). In this case, the chain of custody will be established as it normally would be – through witness testimony at trial. MPD officers will testify about how they obtained and handled the evidence until they delivered it to the FBI for analysis. FBI witnesses will testify about the procedures they followed to ensure the chain of custody. The defendant will be able at that time to pursue any weaknesses in the chain of custody. Until then, the chain of custody issue is not properly before the court. VIII. The Court does not need to hold an evidentiary hearing on all of the defendant’s claims. The Court does not need to hold an evidentiary hearing to rule on the admissibility of the DNA evidence in this case, because it may base its decision on the findings of other courts and other reliable record material such as scientific literature. See, e.g., Karamychev, 772 A.2d at 812 n.8; Porter II, 618 A.2d at 635; Jones, 548 A.2d at 41. See also, e.g., People v. Brown, 40 Cal. 3d 512, 530 (1985) (endorsing practice of looking to other states' precedents in determining whether technique 100/ This case does not involve mitochondrial DNA. 147 is generally accepted), rev'd on other grounds, 479 U.S. 538 (1989); Hadden v. State, 690 So. 2d 573, 578 (Fla. 1997) (same); Owens, 725 N.Y.S. 2d at 182 ("Novel scientific evidence may be admitted without any hearing at all by the trial court."). "Widespread appellate endorsement of a scientific technique should ordinarily end the need for case-by-case adjudication in the trial courts." Grant, 32 Conn. L. Rptr 30, 2002 WL 853627, at *6. As the Connecticut trial court explained, Scientific research is a national – or global – phenomenon, and there is no need for a court in one state to disregard the judicial decisions of other jurisdictions. . . . While a single appellate decision in one state does not automatically settle the issue, at some point the combined weight of judicial and scientific opinion achieves the critical mass of persuasive authority. With respect to STR evidence – at least in the absence of evidence reflecting a change in scientific consensus – that point has now been reached. There is overwhelming evidence that the STR technique has gained general acceptance. This conclusion should end the court's inquiry, and the conclusions derived from that methodology should be held admissible. Id. (internal citations omitted). As we have demonstrated, there is no need for the Court to receive testimony on the general acceptance of the PCR/STR method of DNA typing, or on use of the product rule to calculate random match probabilities. Instead, the Court may properly conclude that those methods are generally accepted based on the overwhelming legal and scientific consensus concerning their admissibility. Similarly, although the Court should hold that the Frye inquiry does not extend to the particular kits used to perform the PCR/STR analysis, we encourage the Court to make a fuller record by ruling, in the alternative, that the Profiler Plus/Cofiler/310 Analyzer system is also generally accepted, as demonstrated by the cases and scientific authorities cited. “The government has amply demonstrated by a preponderance of the evidence that the proffered DNA evidence is generally accepted in the relevant scientific community and is therefore admissible under Frye. None of the issues raised or 148 evidence presented by the defense generated sufficient controversy to preclude the admission of DNA.” Roberts /Veney Order at 9. Further, nothing presented by the defendant has demonstrated that the probability formulas endorsed by the NRC are no longer generally accepted in the relevant scientific community, or that those formulas should not have been used in this case. The Court should therefore admit the evidence and the accompanying probability statistics. We note that all of the above issues have been ruled upon by three judges of this court within the past three years. The only issue that has not yet been addressed in this jurisdiction is the cold-hit issue, but as we have indicated there is no scientific issue to be resolved. However, should the court desire additional information regarding that issue, with sufficient notice to present the necessary experts, the United States will be prepared to address that issue at pre-trial hearings. In conclusion, there is ample scientific support for the admission of DNA evidence in this case. It is therefore appropriate to submit the evidence to the jury for its consideration, subject to the defendant’s ability to test the reliability of the evidence through the traditional means of crossexamination and through his own witnesses. 149 CONCLUSION WHEREFORE, defendant’s motions in limine to exclude DNA evidence should be DENIED. Respectfully submitted, KENNETH L. WAINSTEIN, United States Attorney. VALINDA JONES, D.C. Bar No. 398464, Assistant United States Attorney MICHAEL AMBROSINO Assistant United States Attorneys 555 Fourth Street, N.W. Washington, D.C. 20001 150 CERTIFICATE OF SERVICE I HEREBY CERTIFY that a copy of the foregoing Government’s Superseding Opposition to Defendant’s Motions in limine to Preclude Introduction of DNA Evidence, was served by hand upon counsel for the defendant, Edward Ungvarsky, PDS, 633 Indiana Avenue, N.W., 2nd Fl., Washington, DC 20004, on this 10th day of February, 2016. Michael Ambrosino Assistant United States Attorney 151