FastQC polyA and polyG trimming in RNA-seq

Question

[joonhong kwon]

Hi!

I have raw Fastq data with 151bp paired-end reads, which were trimmed using Trim Galore (v0.6.10). After trimming, I observed overrepresented polyG sequences (purple line) in the FastQC results for Read 2, along with a small presence of polyA sequences (green line). Should I remove both the polyG and polyA sequences, only the polyG sequences, or leave both as they are?

Answer 1

[Mark]

Yes. Trim both before performing any downstream analysis. trimgalore uses cutadapt under the hood, so: https://cutadapt.readthedocs.io/en/v1.8.1/recipes.html#trimming-poly-a-tails

[Istvan Albert]

just make sure the adapter overlap is long enough I'd say over 10 bases

Out of curiosity I checked that 6 base overlap, which typically suffices for artificial adapter sequences would not probably be too short for Gs,

6Gs in a row occur over 6 thousands times even jus the shortest chromosome 22 of the human genome

[joonhong kwon]

Thank you all!