This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Should I remove an overrepresented sequence from one Fastq file from my RNA Seq data?

Hello,

New to RNA Seq analysis and I couldn't find an answer to this question elsewhere. I have an overrepresented sequence in one of my Fastq files (I did pair-end sequencing), which is for one of five drug-treated samples (five control samples, same five drug-treated). This sequence makes up 0.107% of total reads, flagging an amber warning. This sequence was not overrepresented in any of the other control or drug-treated samples, so I'm unsure if this is contamination or something about this sample in particular (the FASTQC report did not identify the source of the sequence = 'No Hit'). I blasted the sequence and the hit was an inflammatory mediator, which isn't totally unrelated to what the research question is addressing.

Should I remove this sequence? Or is it likely this is just biological variance in this one sample responding more profoundly to the treatment?

Thanks in advance!

fastqc overrepresented rna-seq sequences
• 2,089 views • link
updated by Dunois • Dunois • written by robeaumont • robeaumont

The only over-represented sequences you'd have to remove are adapter dimers (if there are any).

• 0 views • link
written by Dunois • Dunois

3 answers

I blasted the sequence and the hit was an inflammatory mediator, which isn't totally unrelated to what the research question is addressing.

There is your answer. Sounds like an expected observation.

There is no rule that every category in FastQC needs to be "green" before you can continue with further analysis.

• 0 views • link
written by GenoMax GenoMax

FastQC was designed for examining DNA, so many of the issues it flags are not problems in RNASeq. An overrepresented sequence taking up a fraction of a percent is not a problem for you.

• 0 views • link
written by swbarnes2 swbarnes2

The whole purpose of doing RNA seq is to measure the changes in gene expression. Or, as you call it, to find over-represented signals (and under-represented). It sounds like you got exactly what you should have, so it is now a matter of properly quantifying your data.

• 0 views • link
written by Mensur Dlakic Mensur Dlakic

Log in to answer this question.

More posts like this