insert size option in paired end

[gupta5...@gmail.com]

Hi Alex

Thanks for your previous reply. I still have to digest most of your answers looking at them more carefully.

I have a question here about insert size option in STAR. Do we have any option in STAR which explicitly mentions the insert size input. For example Tophat has an option -r and then we specify what is the insert size.
Can something of this sort be done in STAR.

If not then how does STAR knows about the paired end data insert size and starts alignment. I mean insert size could be ranging from 200 -400(Encode Caltech GM12878).

Hope to hear from you soon

REgards
VARUN

[Alexander Dobin]

Hi Varun,

STAR does not need such a parameter. When alignments are done to the genome (and not transcriptome), the un-sequenced portion of the insert between the mates may contain a gap (junction) of a much larger length than the inset itself (at least for mammalian genomes) - so the insert size would not be very useful for mapping. My suspicion is that TopHat only uses this parameter when mapping to the transcriptome - see this discussion on SEQanswers:

http://seqanswers.com/forums/showthread.php?t=16472&page=2&highlight=tophat+insert+size

Cheers

Alex