Is it necessary to trim adaptor sequences?

Question

[leekun4]

Hi,

I finished mapping with STAR and counting reads using featureCounts. I noticed that I'm getting low alignment reads (~58%, ~20m reads) for each sample. From my understanding, published papers had higher than 80%. According to the FastQC report, per base sequence content did not pass but the per-base sequence quality was good. If I trim the adopter sequences, I'm afraid that it's going to be too short for alignment (I used SE50).

Any thoughts on this? I really appreciate your help.

[JC]

FastQC report should tell you if you have adapters and if trimming is needed. Also, SE50 is already too short, maybe that is one reason your mapping is low. Are you aligning to a know reference? same species?

[leekun4]

I aligned to a Mus_musculus.GRCm38.98 genome and it's not the same species (I used CD-1).

[JC]

Aligning to a different species can also lead to a minor mapping efficiency

[swbarnes2]

Short reads might cause multi-mapping, but not failure to map. 50 is plenty long, and mots multi-mappers are aligning to truly repetitive regions, so longer reads will only do so much.

[cpad0112]

Trimming adapter sequences - is it necessary?

https://biostar.usegalaxy.org/p/26923/

http://seqanswers.com/forums/showthread.php?t=71558

[leekun4]

Thanks for sharing the links. These posts help a lot.

Answer 1

[swbarnes2]

I'm afraid that it's going to be too short for alignment (I used SE50).

If trimming adapters makes a read that aligns not align...it should not align.

Trimming shouldn't hurt...but it might not help. In my experience with STAR, STAR had no problem soft trimming adapter sequences to make them align.

[leekun4]

Thanks for your response. If you don't mind me asking, which software do you use to do your RNA-seq? Thanks for your help.