Short Insert size and aggressive trimming

Question

[Sharon]

Hi All

When I have short insert size where it is possible that the adapters run into the reads. I trimmed the adapters using cutadapt/trimgalore. FastQC reports regarding adapter contents show no adapters. Is this sufficient or I do need to do more aggressive adapter trimming? Keeping in mind that we still have around 70% alignment with Tophat or 75% mapping rate with Salmon. My professor suggest that I still have adapters in the reads. Is it possible if FASTQC report no adapters. So how can I know how much more trimming is yet needed?

Thanks

[GenoMax]

I was helping someone else with trimming this morning and you may want to try the suggestions there with your data (C: Adapter percentages in the reads )

You will never be able to map 100% of the reads. Best you can do is reach low 90's. At this point you can either move to the next step (using data from TopHat or Salmon) or try a different aligner (STAR or BBMap to see if you get better results). In any case do not expect to get 100% read mapping.

BTW: Many aligners will soft-clip adapters/unmapped sequences so some choose not to trim initial data (I don't).

[GenoMax]

My professor suggest that I still have adapters in the reads.

Just because (s)he feels so? Post FastQC images from before/after if you want to get second opinion here.

[Sharon]

Thank you genomax.Because of the low alignment he thinks. This is after trimming ! https://ibb.co/m79Gxb

[Gabriel R.]

In ancient DNA, we have very short fragment size. You can also try leeHom: https://grenaud.github.io/leeHom/

[Sharon]

Thanks Gabriel. Will give it a try :)

[Gabriel R.]

let me know if it works and need further help :-)

[Sharon]

Sure Gabriel, thanks so much :)