How to gain the most from LongTrace
Posted by: nucleics
Date: September 29, 2006 11:57PM
Advice on how to gain the greatest read length improvement from LongTrace:
1. LongTrace makes good traces better, not bad traces good. LongTrace isn't magic and it can't increase the read length of traces where the signal is just not there. For example, if you have a trace that only got 500 bases before the sequencing signal fell to the level of the detector noise then there is no way for any software to increase the read length to 800 bases - any attempt to do so would just turn machine noise into random peaks.
A good rule of thumb is if you can't see real peaks at the end of the trace then the read length increase offered by LongTrace will be quite limited.
2. LongTrace can only increase the read length of traces when the trace was run long enough to collect the late signal. This means that if you (or your sequencing facility) are stopping the run early (say around base 800) then LongTrace will not be able to supply much of an increase (ie LongTrace can only improve the bases that are collected).
One way to see if you are stopping your runs too early for LongTrace is have a look at the end of the traces after processing them through LongTrace. If the sequencing peaks look good right up to the the end of the trace then this is an indication that you have stopped the run too soon.
A good rule of thumb is to collect around 900 bases with the 36cm capillary arrays, and around 1100 bases with the 50cm arrays.
3. Re-optimise your sequencing reactions. One of the biggest limitations we see with peoples traces is they have optimised their sequencing protocols for short runs. With LongTrace you need good signal late in the trace - many traces that we see just don't have any peak signal after base 800.
It is worthwhile running some of your traces through our free online QualTrace program (http://www.nucleics.com/qualtrace-dna-sequencing-demo/). If your traces are mostly characterized as "excellent" or "good" then your current sequencing protocol is fine, if they are most "end weak" then you should look at modifying your protocol to increase the signal at then end of the trace.