Q: What parameters should I need to tweak for viral ChipSeq

[Rohit-Satyam]

I came across several posts that the MACS is not fine-tuned for Viral ChipSeq (eg SARS-CoV2) (See Post1, Post2. What parameters do you think needs to be adjusted to get reliable peaks?

[taoliu]

@Rohit-Satyam Couple of suggestions for small genome peak calling: 1. usually you need to downsample your data. MACS uses the p-values to measure significancy and a p-value is too sensitive when the sample size is overly large. You may get a significant p-value even if the fold change is small. A 50-100 fold genome coverage may already be good enough. ( fold coverage ~= read_lengthnumber_of_reads/genome_size, or number_of_reads ~= fold_coveragegenome_size/read_length ). 2. Assign a reasonable 'effective genome size' for callpeak -g option, if you know the genome size of the virus. 3. As the Post2 mentioned, try to use the fine-tuning steps described in this document: https://macs3-project.github.io/MACS/docs/Advanced_Step-by-step_Peak_Calling.html since it's more flexible.