Can be merged biological replicates?

Question

[star]

Can we merge/ combine biological replicates from same sample together in ChIP-seq / ATAC-seq data when we want to do peak calling (or like RNA seq we should process each replicate separately)? because I am seeing in some papers, the data were merged for peak calling and some where else processed separately!

Thanks in advance!

[Sej Modha]

I'd perform peak calling on individual samples and then use bedtools intersect or mergePeaks from Homer package to assess the overlap among replicates.

[Benn]

I have used that approach before for ATACseq. So first call peaks for all bam files at once, and then quantify how many reads were mapped to the peak regions for each sample (bam file) for statistical analysis in edgeR.

[1769mkc]

" and then quantify how many reads were mapped to the peak regions for each sample (bam file)" how did you take out the count? by using multibamsummary? or

Answer 1

[ZZzzzzhong]

For the experiment with several replicates, it is recommended to concatenate several ChIP-seq treatment files into a single file. To do this, under Unix/Mac or Cygwin (for windows OS), type:

$ cat replicate1.bed replicate2.bed replicate3.bed > all_replicates.bed

For BAM or SAM files, samtools can be used to combine replicates.

at MACS2 Manual(https://github.com/taoliu/MACS) recommend to merge biological replicates.But you can also call peak separately then use overlapping peaks

[ATpoint]

If you merge replicates prior to sequencing, it might be useful to downsample the data to have equal readcounts. Otherwise, the peak calling is dominated by the sample with the highest counts. See this post on BioC for details.

Also @ZZzzzzhong, if you copy statements from a website (the first sentence of your answer), please indicate it by using the Blockquote option in the formatting bar and putting in a link to the reference directly behind it.