fastqc - overrepresented sequences

[PFS]

I have run a FASTQC analysis and found out that there are hundreds of over-represented sequences in my datasets. Some of these, are Illumina PCR primers, some are single end adapters.

Does it mean I have some contamination going on? What can I do about it? Do I simply remove the primer and adapter sequences or what else?

Thanks in advance.

[BAMseek]

If your data is microRNA or something where the sequence length is less than the read length, then you might end up reading into adapter or primer sequences. This would possibly show up as over-represented k-mers on the 3' end of the sequences. In this case you might want to trim the adapters before alignment.

[simonandrews]

If they're in the overrepresented sequences then it probably means that your library was contaminated with primer dimers. If you read through into adapter you'd get slightly different sequences so they would appear in the Kmer plot rather than the overrepresented sequences.

There's not much you can do about the dimers for the data you've already collected. They're easy enough to filter out if you want to remove them, but they probably won't map to whichever genome you're using anyway.

You can normally spot this kind of contamination by doing a BioAnalyser run on your library before sequencing. Others here are better qualified than me to advise on how you might avoid them in the first place.

[DZhang]

Another source contributing to the over-represented sequences is reads from rRNA genes. Even with rRNA removal, oftentimes you still have high-level rRNA reads.

Douglas

https://www.contigexpress.com