de novo transcriptome assembly, read length?

Question

[nikelle.petrillo]

Hello

I have quality filtered my reads and I have been advised to throw out any reads with a length of 45 nucleotides of less. I used fastqc in galaxy and I can see that pre filtering, my "sequence length" is 100. Post filtering, my "sequence length" is still 100. Is this the right parameter to be looking at if I want to figure out what reads should be thrown out?

Thanks for your help.

Nikelle

[GenoMax]

If the read length is still 100 post-filter then a) either your data has no contamination of adapters etc or b) the filtering/trimming did not work as expected. But that would be the right parameter to look at.

[pld]

Is this before or after quality trimming?

[Michael]

I agree with genomax that the data might just be of good quality and not necessarily reason for concern. When trimming some recent Nextseq data using trimmomatic, we have almost seen no adapter contamination and lost much less than 1% of the whole reads due to qual-trimming, such that we decided we don't need trimming at all, but that was not de-novo assembly. If it is still relevant for your project, you could post the filter parameters and an example of the fastqc output, it still looks like rather good than bad news to me.