BNFO 601
Usman Roshan

• Perl progamming language (and some Unix basics)
– Unix basics
– Intro Perl exercises
– Dynamic programming and Viterbi algorithm in Perl
• Sequence analysis
– Algorithms for exact and heuristic pairwise alignment
– Hidden Markov models
– BLAST
– Program parameter training for alignment
• Population structure identification using genome-wide SNP data (time permitting)

• Grade: 50% mid-term and 50% final exam
• Exams will cover Perl and bioinformatics algorithms
• Recommended Texts:
– Introduction to Bioinformatics Algorithms by Pavel
Pevzner
– Biological sequence analysis by Durbin et. al.
– Introduction to Bioinformatics by Arthur Lesk
– Beginning Perl for Bioinformatics by James Tisdall

-3 mil yrs
-2 mil yrs
AAGGCTT
AAGACTT
T_GACTT
-1 mil yrs
_GGGCTT TAGACCTT A_CACTT
GGCTT
(Mouse)
TAGGCCTT
(Human)
TAGCCCTTA
(Monkey)
A ACACTTC
(Lion)
C ACCTT
(Cat)
CTT
(Cat) today

Representing DNA in a format manipulatable by computers
• DNA is a double-helix molecule made up of four nucleotides:
– Adenosine (A)
– Cytosine (C)
– Thymine (T)
– Guanine (G)
• Since A (adenosine) always pairs with
T (thymine) and C (cytosine) always pairs with G (guanine) knowing only one side of the ladder is enough
• We represent DNA as a sequence of letters where each letter could be
A,C,G, or T.
• For example, for the helix shown here we would represent this as CAGT.

Proteins are chains of amino acids. There are twenty different amino acids that chain in different ways to form different proteins.
For example,
FLLVALCCRFGH
(this is how we could store it in a file)
This sequence of amino acids folds to form a 3-D structure

• The protein folding problem is to determine the 3-D protein structure from the sequence.
• Experimental techniques are very expensive.
• Computational are cheap but difficult to solve.
• By comparing sequences we can deduce the evolutionary conserved portions which are also functional (most of the time).

• Primary structure: sequence of amino acids.
• Secondary structure: parts of the chain organizes itself into alpha helices, beta sheets, and coils. Helices and sheets are usually evolutionarily conserved and can aid sequence alignment.
• Tertiary structure: 3-D structure of entire chain
• Quaternary structure: Complex of several chains

• DNA can be represented as strings consisting of four letters: A, C, G, and T.
They could be very long, e.g. thousands and even millions of letters
• Proteins are also represented as strings of 20 letters (each letter is an amino acid).
Their 3-D structure determines the function to a large extent.

• How to align two sequences?

• How to align two sequences?
• We use dynamic programming
• Treat DNA sequences as strings over the alphabet {A, C, G, T}

Define V(i,j) to be the optimal pairwise alignment score between S
1..i
and T
1..j
(|S|=m, |T|=n)

Define V(i,j) to be the optimal pairwise alignment score between S
1..i
and T
1..j
(|S|=m, |T|=n)
Time and space complexity is O(mn)

How do we understand this dynamic programming algorithm?
• Let’s first look at some example alignments
• Let’s look at gaps. How do we know where to insert gaps
• Let’s look at the structure of an optimal alignment of two sequences x and y and how it relates optimal alignments of subsequences of x and y

Animation slides by Elizabeth Thomas in
Cold Spring Harbor Labs (CSHL) http://meetings.cshl.org/tgac/tgac/flash/DynamicProgramming.swf

• Recall that proteins have 3-D structure.

Alignment of thioredoxins from human and fly taken from the
Wikipedia website. This protein is found in nearly all organisms and is essential for mammals.
PDB ids are 3TRX and 1XWC.

Unaligned proteins.
2bbm and 1top are proteins from fly and chicken respectively.
Computer generated aligned proteins

• We can produce high quality manual alignments by hand if the structure is available.
• These alignments can then serve as a benchmark to train gap parameters so that the alignment program produces correct alignments.

• Protein alignment benchmarks
– BAliBASE, SABMARK, PREFAB,
HOMSTRAD are frequently used in studies for protein alignment.
– Proteins benchmarks are generally large and have been in the research community for sometime now.
– BAliBASE 3.0

Biologically realistic scoring matrices
• PAM and BLOSUM are most popular
• PAM was developed by Margaret Dayhoff and co-workers in 1978 by examining
1572 mutations between 71 families of closely related proteins
• BLOSUM is more recent and computed from blocks of sequences with sufficient similarity

• We need to compute the probability transition matrix M which defines the probability of amino acid i converting to j
• Examine a set of closely related sequences which are easy to align---for PAM 1572 mutations between 71 families
• Compute probabilities of change and background probabilities by simple counting

• Basics of Unix
• Perl programming
– Basics
– Exercises
– Dynamic programming alignment solution in
Perl