Quality control steps before alignment?

[kjaja]

Hi,

Are any quality control steps that I need do before aligning the reads to the reference genome using the fastq files?

[upendra_35]

There are innumerable number of threads related to this top. Try searching......

[Heisman]

You can look into using FastQC to see if the data set is worth analyzing. Other than that, most QC steps can be implemented after aligning (ie, tossing out reads with high error rates or low mapping qualities, for example).

[simonandrews]

There are some QC steps which would need to be done before aligning. In particular the trimming of any adapter sequences found by QC checks would need to be done early on.

I'd also stress that even if the QC is awful I'd still go ahead and analyse it anyway. We've found biologically interesting results from samples with 97% duplication. When you've got 40 million reads, even if you throw 97% away you still have quite a few left :-)

We view QC as a means to allow us to better evaluate the results of subsequent analysis and to help diagnose problems, rather than a filter to let us throw samples away.