ChIP-Seq Histone modification time span study

Question

[Sudhir Jadhao]

I am analysing ChIP-seq data for histone protein.

The study is timeline study. It aims to study the binding pattern of histone at different time points(0hr, 2hr, 4hr, 5hr, 6hr). We have two strains namely virulent and avirulent for the time points (2hr, 4hr, 5hr, 6hr) and 0hr sample is common.

I have a few questions about analysis listed below:

1. I am doing peak calling with help of MACS2. What are the parameters to be consider for peak calling?
2. I don't have Input and mock, should I use one of my sample as control during macs2 peak calling or I do peak calling step without control
3. What should be the minimum number of peaks discover in histone study?

Answer 1

[Manvendra Singh]

1) in -f you specify if your allignment is in bam format or sam format, -g specify the genome which you are mapping (e.g. hs for Human)

-n specify the names of output, -q is FDR threshold, I also give -B, where I also get bedGraph file of peaks, its convenient for some statistical tests. -t is treat, oc is control

-f <SAM/BAM> -g <hs> -n <names> -q <0.05> -B

2) I would suggest you to always call peaks taking Input as control, for Histones and mock as well., Later compare Histones with mock to see the relative enrichment against mock.

3) It depends, See how many you get first.

[Sudhir Jadhao]

Manu, Thanks for reply

In some research paper suggested that if you are studying time span study then call peak without using control