interpretation of fastQC result

[liaojinyue]

Dear All,

I'm just got my RNA-seq result back but I'm new to analysis. I decided to try it out following the tuxedo protocol. After alignment with Tophat2 following quality trimming, I got relatively low mapping rate (~50%-60% ). I'm not sure what's the cause. I'm thinking about cleaning the data such as trimming the adapter before doing the alignment. I can tell the fastQC report doesn't looks so good but have no idea where to start. Can anyone have a look at the report and give me some suggestions? The libraries are prepared using integenX kit after rRNA depletion and we sequenced them by PE100. Thanks.

Jason

[Brian Bushnell]

That looks pretty bad. You should probably examine the unmapped reads to see what they have in common; maybe you can BLAST them to figure out what kind of artifacts are present. Once you know, you can easily filter or trim the offending sequences with BBDuk.

Or you can just start with adapter trimming, assuming TruSeq adapters were used, and see if that solves the problem. If you post the read insert size distribution it will be more obvious whether or not the problem is caused by adapters.