Per base sequence content: "A" underrepresented

Question

[Wöps]

Dear all,

my RNA-seq reads seem to be slighly biased to show more 'T' than 'A' (see image). This is unexpected for me. Does anyone have a suggestion why that might be (and how to proceed)?

The data are paired-end, sequenced with Illumina HiSeq 2500. (The read direction is reverse.)

thanks for your help!

[benformatics]

Your image doesn't work but also something to think about is that the most RNA-seq technologies sequence the cDNA (reverse transcribed copy) of your mRNA molecules. Thus assuming the method is strand-specific the poly-A tails will often look like stretches of T's; potentially explaining your results.

[ATpoint]

Fixed the link to the image. Please use the image button and paste in the full link incl. the suffix (e.g. .png) into the field that pops up:

[Wöps]

Thank you! Will do next time.

[russhh]

Have you trimmed off poly-A sections?

[Wöps]

I hadn't trimmed poly-A sections so far. I now quickly tried running afterQC (https://github.com/OpenGene/AfterQC) to remove any reads with polyX above length 20. It does not seem to change the situation though:

[ATpoint]

As genomax says, what you see is normal. Just proceed with your analysis.

Answer 1

[GenoMax]

This is a fairly classic example of RNAseq libraries. See this blog post from authors of FastQC for more details.