Failed FastQC Per sequence GC content of MeDIP Input sample

Question

[Gary]

Hi,

We have two Input samples (not IP samples) of MeDIP-Seq analyzed by FastQC. One of Input samples has warning Per sequence GC content (fig1), and the other has failed Per sequence GC content (fig2). Our human mesenchymal stem cells (MSCs) data are 100 bp paired-end reads sequenced by Illumina HiSeq 2000. Do you know what possible problems with our Input samples? Many thanks.

Best,
Gary

fig1_T8L Input

fig2_T7E Input

Answer 1

[Matt Shirley]

MeDIP-seq is going to give you methylated DNA, which is mostly CpG islands, so is it really surprising that you might have an abnormal GC content after this selection? I wouldn't assess GC content as a QC metric for an enrichment method like this unless I was certain what kind of distribution to expect. FastQC is telling you "failed" based on an expectation of a poisson sampling from a uniform distribution of GATC, which you clearly will not have in these libraries.

[Gary]

Many thanks for your information. However, the two samples are Input without 5mC antibody enrichment.

[Matt Shirley]

Sorry, I didn't fully read your question.

[Gary]

It's OK. I appreciate your help, and any other suggestion is very welcome.