What is the best method ChIPseq differential binding analysis?

Question

[biopat]

Hi all,

I want to quantitatively compare ChIP-seq data!

Experimental condition: - same background cell line but samples differ (WT, KO1) - same antibody was used to pull down transcription factor (not histone) - identical library preparation and sequencing parameters

My goal is to perform a differential binding analysis between the two samples (I have multiple biological replicates per condition). Considering I am looking at a transcription factor and have multiple biological replicates, which differential binding analysis is the most reliable for whole genome differential binding analysis?

I have checked out two review papers: 1. Shiqi Tu et al. An introduction to computational tools for differential binding analysis with ChIP-seq data 2. Sebastian Steinhauser et al. A comprehensive comparison of tools for differential ChIP-seq analysis

Based on their flow charts I have following options: -diffReps-nb -MultiGPS -PePr -ChIPDiff -ODIN -THOR

I am not sure if any of these are what I am looking for, so any help would be appreciated!

Thank you!

[Carlo Yague]

I can't tell you about the other tools, but ODIN doesn't take replicates into account, so you can cross this one from your list. THOR (from the same group) uses replicates and is a better option, especially if you have input controls.

[zhenyisong]

Beside @Rory's suggestion, I would suggest you to pay attention to this paper (The Overlooked Fact: Fundamental Need for Spike-In Control for Virtually All Genome-Wide Analyses) about the wet procedure. Sebastian's work argued that there were few overlaps between different pipelines.

Answer 1

[Rory Stark]

Disclosure: I am the author of DiffBind, a differential binding analysis tool.

Interestingly, the tools you are considering are not the ones most often used to derive results in the published literature. MACS + DiffBind is by far the most widely used combo for experiments with multiple biological replicates, being the tool of choice in more than 200 peer-reviewed publications reporting biological results. This tool is recommended explicitly by the Steinhauser paper. The Shiqi Tu paper does not evaluate DiffBind directly, but recommends the methods it encapsulates (called the edgeR and DESeq methods in the article). DiffBind is well-supported, feature-rich, and includes a clear, helpful tutorial (vignette).

Another tool that is worth a look is csaw, which lets you skip the peak-calling step. Both DiffBind and csaw are part of R/Bioconductor and integrate with other Bioconductor tools.

Regards-

Rory

[biopat]

Thank you very much! I appreciate your input!

[Researcher]

Hi all,

I want to quantitatively compare ChIP-seq data between different subtypes of a cancer.

My experimental design is as follows: I have multiple samples enriched for 3 different histone marks from 2 subtypes of a cancer and there are no replicates for them. The sample structure somewhat looks like this: Subtype 1 Subtype 2 Sample_A_K4me3 Sample_1_K4me3
Sample_A_K27me3 Sample_1_K27me3 Sample_A_K27ac Sample_1_K27ac Sample_A_input Sample_1_input

Sample_B_K4me3 Sample_2_K4me3
Sample_B_K27me3 Sample_2_K27me3 Sample_B_K27ac Sample_2_K27ac Sample_B_input Sample_2_input

Sample_C_K4me3 Sample_3_K4me3
Sample_C_K27me3 Sample_3_K27me3 Sample_C_K27ac Sample_3_K27ac Sample_C_input Sample_3_input

Since, I am interested to call differential binding sites between the subtypes of a cancer, Can I use these samples (like samples A,B and C) which are enriched for same histone mark (say K4me3) from same subtype as a surrogate to replicates and compare it with the similar set (1,2 and 3) from other subtype?

If this is ok which tool is recommended to use in order to compare these unpaired samples and call differential binding sites between both the subtypes.

[Jingyue]

Hi, Rory, Thanks for the software, I'm wondering do you recommended overlap the differential peaks between softerware, like between DiffBind and Csaw? Will this give a better/accurate differential results?