Extremely high adapter content and low alignment rate in RNAseq

Question

[grayapply2009]

Hi guys,

I recently encountered some RNAseq data with extremely high adapter content (over 90% of reads with about 50% adapter contents) and low alignment rate (<2%, GRCH38).

Are these reads contamination (not from human)? Should I discard the reads with high adapter contents? Is there a way to rescue these data?

The pic shows the alignment after trimming adapters using trimgalore with default settings.

Answer 1

[swbarnes2]

If it's adapter, it's adapter, not contamination. You might be able to rescue the reads by trimming the adapter

[grayapply2009]

They don't look like adapters. They are 50bp long. I trimmed them then the reads are too short to align.

[swbarnes2]

Whelp, that's all you can do. The library prep was no good, your fragments are so small that you are reading through to adapter