Calculate enrichment values of Chip-seq peaks for hierarchical clustering

Question

[florian.noack]

Hi everyone, I have Chip-seq samples + input controls (triplicate) for three different cell types. I want to study the changes of the Chip-seq samples between the three different cell types. I called already peaks using MACS2 and generated a global peak file (if a genomic regions pops up in at least one of the three cell types its in this file). No I wondering how I can calculate for each sample (should I merge the replicates, they show a very high pearson correlation) a enrichment value which considers also the input control. Of course the enrichment value should be also normalized by the total readcount to be comparable between the different samples. I would like to feed this enrichment value into pheatmap for hierarchical clustering.

Thanks a lot.

Flo

Answer 1

[igor]

You should look into DiffBind or csaw. There are a few tools for differential binding that will merge, quantify, and normalize peaks across multiple samples/groups.

• https://bioconductor.org/packages/release/bioc/html/DiffBind.html
• https://bioconductor.org/packages/release/bioc/html/csaw.html