Supplemental materials.
To see if one could use this approach to identify viral isolates, we conducted a
set of studies in which the identities of the viral strains were unknown to those
performing the experiments and data analysis. Four anonymous strains of human
rhinoviruses were obtained from our collaborators. Before constructing their
sequence motif maps, the nearly full length dsDNA was generated by reverse
transcription followed by long range PCR with a pair of universal PCR primers for
all strains. PCR forward primer is an 18bp oligo (5’TCCTCCGGCCCCTGAATG),
which is conserved for all the strains and the reversed primer
(5’Cy3ATTCTAGAGGCCGAGGCGGCCGACATG-d(T)30) was labeled with Cy3
to indicate the orientation of the DNA molecules. In these experiments, the
nicking endonuclease Nb.Bsm I (GCATTC) was used and the nicking sites were
labeled with Tamra-ddCTP. After labeling, the samples were purified with S400
spin column (Amersham, UK) to remove the labeled PCR primers. The predicted
maps for all four strains are shown in Figure S1. Figure S2 shows some typical
false color two-channel composite images of the stretched DNA contours (YOYO
in blue) and the Nb.Bsm I sites (Tamra-ddCTP in green). On average, the
labeling efficiency is about 50%. The DNA molecules that stretched over 2.0
microns in length with one end green label and at least one green internal label
were chosen to construct sequence motif maps shown in Figure S3. About 100200 DNA molecules were used. Clearly, three peaks at 0kb, 1.6 kb and 3.69 kb
are observed in sample 1. The two internal peaks of the map match well with the
predicted nicking sites of HRV-15 virus at 1.57kb and 3.62kb respectively. Only
one internal peak was identified in sample 2 at 3.52 kb, which is consistent with
the map of HRV28. Sample 3 was confirmed to be HRV36 virus. Interestingly,
the HRV15 and HRV36 would not be distinguished if no end label was used to
indicate the orientation of the DNA molecules. Sample 4 is a little bit problematic
since some of DNA molecules have two end labels, which makes it impossible to
discern the orientation of the DNA molecules. The map of sample 4 was
generated disregard of the end label. Two pairs of mirrored peaks were observed
at 0,6.7kb and 1.53kb, 5.27kb, which correspond to the predicted map of HRV73
virus. The minimal distance of 1.6kb was detected between two closest labels,
indicating the resolution of 1.6 kb can be achieved.
Figure S1. Sequence motif maps (Nb.Bsm I) of human rhinovirus 15, 28, 36 and 73.
1.57kb
GCATTC
3.62kb
GCATTC
6.82 kb HRV_15 DNA
3.63kb
GCATTC
6.78 kb HRV_28 DNA
3.84kb
GCATTC
5.52kb
GCATTC
6.78 kb HRV_36 DNA
1.65kb
6.76kb
GCATTC
GCATTC
6.78 kb HRV_73 DNA
Figure S1. The predicted Nb.Bsm I (GCATTC) map of human rhinovirus 15, 28,
36, and 73.
Figure S2. The intensity scaled composite two-false-color images.
Sample 1
5 micron
Sample 2
5 micron
Sample 3
Sample 4
5 micron
5 micron
Figure S2. The intensity scaled composite two-false-color images. In the large
field of the intensity scaled composite image, numerous molecules are found.
Some molecules were fully labeled (red arrows) and some were partially labeled
(yellow arrows).
Figure S3. The constructed sequence motif maps.
70
80
Sample 2
Experimental data
Gaussian fitting
Sample 1
60
50
Frequency
Frequency
Experimental data
Gaussian fitting
60
40
3.69kb
1.62kb
40
3.52kb
30
20
20
10
0
0
0
1
2
3
4
5
6
0
7
1
2
DNA Length (kb)
70
4
5
6
7
80
Sample 3
Sample 4
Experimental dat
Gaussian fitting
60
Experimental data
60
Frequency
50
Frequency
3
DNA Length (kb)
40
3.86kb
30
5.76kb
1.53kb
5.27kb
40
20
20
10
0
0
0
1
2
3
4
DNA length (kb)
5
6
7
0
1
2
3
4
5
6
DNA Length (kb)
Figure S3. The constructed sequence motif maps. The sequence motif maps in
the graph were obtained by analyzing 100-200 DNA molecules. The solid line is
the Gaussian curve fitting and the peaks correspond well to the predicted
locations of the sequence motif.
7