Mapping and filtering treatment and control ChIP-Seq data for MACS2

Question

[wardahjassiem97]

Hi

I am following these steps to prepare my treatment and control chipseq data for peak calling with MACS2; mapping with bowtie2, sorting, indexing and filtering for uniquely mapped reads with sambamba.

The instructions are from this site: https://github.com/hbctraining/Intro-to-ChIPseq/blob/master/lessons/03_align_and_filtering.md

Do I repeat these steps (mapping, filtering for uniquely mapped reads etc) on the control/input files as well as the treatment files?

Thank you

Answer 1

[LChart]

Yes. The control is to provide a background; and thus should be treated the same way as the treatment files.

[wardahjassiem97]

Thank you!

[ATpoint]

Be careful though. Often ChIP-seq controls, for example IgG or chromatin input are so undersequenced that they're of little use with most of the genome not being covered. Since peak callers such as macs will downsample the larger to the smaller dataset it can well be that the downsampling throws away so many reads from the treatment that your peak calling suffers a lot and you might be better off without a control.