Trim Illumina reads?

[sapearl]

Hello,

I was wondering if anybody knows whether or the degree to which removing the low quality ends of Illumina reads (100x2 paired end) will improve denovo assembly? When asking my colleagues as to whether I should "clean" my data by first quality trimming it, I've gotten mixed responses and I haven't come across any papers addressing this. Thoughts?


Thanks!

[DZhang]

Read clean-up is the standard initial step in my de novo assembly pipeline. The assembly process is prone to low-quality data and sequencing errors.

[lexa]

I also would recommend read trimming to get rid off poor quality bases that mess up the de novo assembly. in our lab, it is the standard procedure before we do any de novo steps.

[eslondon]

I have applied it systematically to several de novo assemblies and it actually made the assemblies worse. I think this will very much depend on the specific tool used for de novo assembly. Best to try it on your own data, with your own tool and compare results.

When you compare results make sure that you not only compare N50, but also quality, e.g. alignment of other known sequences to your genome, etc.