Normalization of ChIP-Seq count matrix ??

Question

[Ankit]

Hi everyone,

I created a count matrix for ChIP-seq data. Basically I counted the reads present inside peak coordinates and made a matrix. Now I have two questions:

Q1. Do I need to normalise count matrix for library size in order to compare among replicates? If yes, can I use similar strategy like RNA-Seq count normalisation implemented in DESeq2/ edgeR?

Q2. Do I need to take care of normalisation by input counts? I think not because peaks are called on the background of input. So Ican use only direct matrix after library size normalisation.

Please suggest something

Thank you

Answer 1

[ATpoint]

The csaw book discusses ChIP-seq normalization extensively. http://bioconductor.org/books/3.13/csawBook/chap-norm.html