|
Epicentre Forum 1 (1)
Scott L. Sutter and David L. Steffens, LI-COR, Inc., P.O. Box 4425,
4421 Superior St., Lincoln, NE 68504.
Recent advances in automated DNA sequencing technology have substantially
increased the amount of accurate sequence information obtained per reaction1.
The LI-COR® Model 4000L infrared fluorescence-based DNA Sequencer
can routinely separate and identify 800-1000 bases in a sequencing reaction.
We developed an effective, simple procedure for sequencing both single-
and double-stranded DNA templates using the Model 4000L Sequencer and
the SequiTherm Long-Read Cycle Sequencing Kit. This procedure can give
accurate sequence data up to 1000 bases.
Figure 1 shows the Model 4000L sequence output for M13mp18 or pBS+ plasmid
DNA control templates sequenced using the SequiTherm Long-Read Kit. DNA
sequencing video images generated by the Model 4000L are similar to autoradiography
images. In the sequencing run shown, these samples were sequenced to >900
bases each with less than 1% error.
 |
Figure 1. LI-COR Model 4000L video print of M13mp18 (left) or
pBS+ (right) DNA sequenced using the SequiTherm™ Long-Read™ Kit. Sequencing
reactions were run with the SequiTherm Long-Read Kit as described
in the kit manual (we found that the standard SequiTherm Kit could
also be used if modified somewhat, however, base reads were shorter).
Primers were 5'-end labeled with an infrared dye, IRD41. The M13mp18
sample was sequenced to 989 bases (0.9% error). The pBS+ sample
was sequenced to 950 bases (0.7% error). |
Table 1 presents the results of multiple reads by the LI-COR Base ImagIR™ image
analysis software of sequence data from M13mp18, pBS+ and 4 double-stranded
plasmids containing inserts of unknown sequences using the SequiTherm
Long-Read Kit. Read lengths were calculated from where the software's
autosequencer stopped calling bases due to low intensity bands, blurry
bands or a high uncertainty level for base calls. The majority of reads
fell into the 700-899 bases range, with most single-stranded DNA samples
giving reads of 800 bases or greater.
Table 1. Summary of Read Lengths
| Sample |
Read Length |
| |
400-499 |
500-599 |
600-699 |
700-799 |
800-899 |
900-999 |
>1000 |
| Single stranded test template (M13mp18) |
0 |
1 |
1 |
4 |
8 |
4 |
3 |
| Double stranded test template (pBS+) |
1 |
1 |
0 |
5 |
2 |
1 |
0 |
| Double stranded with insert |
0 |
4 |
3 |
8 |
9 |
3 |
0 |
| Totals |
1 |
6 |
4 |
17 |
19 |
8 |
3 |
|
Table 2 shows accuracies of computer-analyzed sequences obtained from
the Model 4000L/SequiTherm Long-Read Kit procedure. Comparisons of the
data with the known sequences were done using DNASTAR's MegAlign™ program.
Error rates were calculated by dividing the total number of errors by
the total number of bases in a cumulative manner. Error rates were much
better for single-stranded DNA templates (less than 1.0% error at 800
bases) than for double-stranded samples (2.45% at 700 bases).
|
Table 2. Average Cumulative Reading Errors
| Average % Error |
Read Lengths |
| |
to 300 |
to 400 |
to 500 |
to 550 |
to 600 |
to 650 |
to 700 |
to 750 |
to 800 |
to 850 |
| Single-stranded M13mp18 (n=21) |
0.52 |
0.70 |
0.56 |
0.55 |
0.53 |
0.57 |
0.60 |
0.81 |
0.99 |
1.08 |
| Double-stranded pBS+ (n=10) |
0.87 |
1.08 |
1.54 |
1.63 |
1.84 |
2.13 |
2.45 |
2.91 |
- |
- |
|
Precision of the Model 4000L/SequiTherm Long-Read Kit procedure was
determined for 3 of the 4 double-stranded plasmids with inserts, by calculating
sequence differences between multiple sequence analyses of the same template.
For one plasmid, multiple sequence results showed less than 1% difference
at 700 bases; results for the other 2 plasmids showed less than 1% difference
at 850 bases.
- L. Middendorf, S. Roemer and D. Grone, Improving the Resolution
of Automated DNA Sequencing, Genome Sequencing and Analysis
Conference V, Abstract C20, Hilton Head, SC, (Oct. 23-27, 1993).
|
