Fastqc shows over-represented sequence with no sequence given?

Question

[chapmansm13]

Hi,

I uploaded a sequence file to FastQC and it showed an over-represented sequence so I trimmed it with cut-adapt. I re-ran the trimmed file through fastQC and it shows another over-represented sequence with a different count and percentage saying 'no hit' and not showing the bases of the sequence?

Does anyone know why its not showing the bases of the new hit??

Thanks

Answer 1

[John]

If im not mistaken, the 'No Hit' refers to possible sources for the over represented sequence (i.e., known primer/adapter sequences)
So this simply means that whatever your over represented sequence is, FastQC doesn't know about it in its contaminants list.

This isn't really a problem - just cut-adapt on whatever you think the sequence is. Better yet, paste the sequence into Google to find out what it is :)

In future, you will get a better answer if you screenshot your data/problem :)

[chapmansm13]

Hi, its not actually telling me what the sequence bases are.

[John]

What the... o_o
...well, er, i'd like to say don't worry about it because 5547 of anything is not a lot.
... but since FastQC is clearly not behaving itself, perhaps you should run a different over-represented sequence tool on it.

[chapmansm13]

Ok thanks, Are there any tools you would suggest?

[John]

Sorry I don't know of any other than FastQC, but I'm sure a quick Google will bring something up.

I wrote a module for SeQC that does this, but its a bit technical (you'd need to know Cypher or have Gelphi installed..?) and its very unpolished. If you're willing to wait a month I can help, hahah, but your best bet is to just find another tool like FastQC :)

Answer 2

[MboiTui]

This is the result of not removing sequences with 0 length after trimming (hence why it is not reporting the base pairs). Set the minimum sequence length to more than 0 and it will solve the issue