Hi, I have a problem processing ChiP-Seq data. I have two datasets of H3K9me3 replicates aligned to the reference genome mm39. For peak calling, I used hiddenDomains.
About my problem: replicate A shows peaks on all chromosomes after peak calling, replicate C shows no peaks on Chr12. The .bw files of both replicas show reads on Chr12.
The manual of hiddenDomains says that you can change the parameters min.read.count, max.read.count and min.prob in case of problems. Testing different values of the three parameters (especially setting max.read.count lower or higher) did not fix my problem with replica C.
In hidden domain it is stated that the peaks of each chromosome are called independently. However, when I remove ChrX and ChrY of the .bam files before peak calling, replicate A shows marginal changes, and with replicate C I additionally lose the peaks of Chr4.
To visualize my problem: On the igv snapshot below, replica A is blue, replica C is red. The top lane for each replicate shows the .bw file. The track below is the .bed file I got by peak calling with all the chromosomes, and the bottom track for each replicate is the .bed file I got by peak calling after removing ChrX and ChrY.
Does anyone who is familiar with peak calling with hiddenDomains or has ever had the problem of not being able to get peaks for certain regions despite existing reads in the .bw files and could help me?
Many thanks in advance!