Strange per base quality data - FastQC

[jb1]

Hi,

I have just received a new dataset to work on so I ran it through FastQC.

I got some really strange looking base quality results.

It looks like the vast majority of reads have absolutely terrible quality, and that there are a small subset that look more normal.

This is basically a ChIP-seq data set, sequenced on illumina.

Is the data a bust?

Thanks,

Jon

[fkrueger]

Hi Jon, this does indeed look pretty poor but you might rescue up to a million or so reads by trimming the data (mainly for quality). These reads might still tell you whether the ChIP worked or if there is something you can improve. Several tools will do this trimming for you, e.g. Trimmomatic, Cutadapt, FastX toolkit or Trim Galore.

[NextGenSeq]

Not sure why you did a 100 cycle run for ChIP. We always run SE50 for ChIP-Seq.