illumina aligment ratio reduced when increase insert size

[huan]

We build two different cDNA library for the same sample with different size(196 and 225) for pair end sequencing by Hiseq with the same sequence length 126. Then we map the two set of sequencing data to the genome by tophat. But the result show that the insert size for 225bp cDNA library is only 200bp, so whats the matter of it? How can I deal with it?

Thanks a lot for any possible reply.

[Brian Bushnell]

Why are you using Tophat? Is this RNA data?

Anyway, insert size is pretty difficult to calculate with RNA-seq data; generally, I would expect the reported insert size by mapping to a genome to be longer than the actual insert size. If you are seeing it as shorter, then perhaps the people who made the library undershot the target.

Since your libraries should overlap, you can also get an insert size distribution purely from overlapping, with BBMerge; that will not be biased by genomic factors like like introns.

[huan]

Thanks a lot @Brian Bushnell. You did inspire me. I'll have a try.
Thanks again.