large rearrangements

[dnusol]

Dear list,

I am not sure if this question belongs to the bioinformatics section, but it is indeed a basic question. In the case of a tumor experiment, one could expect large rearrangements, big enough to delete several introns and exons from a gene. Is it possible to detect such large rearrangements using 35 base-long Solexa reads? How big indels could I detect? I guess if the experiment was a pair-end one, that way you could probably detect heterozygous indels. How about homozygous ones? Is it possible to detect this kind of indels by visualizing alignment patterns with, for example, maqview looking for alignment gaps?

Thanks for your help

Dave

[krobison]

With such short reads, it would seem that you have no hope without paired-ends / mate pairs. There are several tools which will do this analysis -- look for BreakDancer & MoDiL & VariationHunter in the software wiki.

[dnusol]

Thanks a lot krobison, so would you say that for detection of indels, pair-end experiments are necessary, and that the longer the reads the better?

[krobison]

At this point I have only book knowledge (well, from papers) but paired-end definitely gives you a lot more information. One way to look at it is the paired-end methods can detect a rearrangement or indel without seeing that event in a read whereas a single read strategy must rely on reads spanning the breakpoint (and any short sequencing read directly spanning a breakpoint will be very challenging to align!). Even if it is possible to obtain & align reads across a breakpoint, the paired strategy is going to give more evidence.

In a tumor setting, you are also dealing with the fact that your sample is highly unlikely to be pure (surrounding tissue, stroma, immune infiltrate). Being more sensitive is critical for this application (which is one I am interested in)

[krobison]

This link will find all the informatics applications in the wiki tagged as being in this space (I just remembed & tagged another one, Pindel)

http://seqanswers.com/wiki/Special:SearchByProperty/Bioinformatics-20method/structural-20rearrangements