Hi, I am quite a novice for NGS analysis. I had a conversion with my colleagues about peak calling for ChIP seq data. I got confused about several comments she mentioned.
comment one "MACS is too old, and no one use it any more", is that true? It seems to me that it is still the most widely used, if not the best, peak caller now. I could still see it being used in the most recent high profile journals.
"Setting the parameter of peak calling is an art", OK, i made this up. She said it is essential to tailor parameters to fit each set of data. I understand that it is important to adjust the parameters based on whether the enrichment is broad (many histone modifications) or narrow (TFs). but she comments seems suggest that you could tailor you parameter as much as I want, as long as you apply the same parameter to the same dataset you are supposed to look at. Are there general rules or principles.
Another is that are there any peak caller optimized for low cell number ChIP seq, say 110^5 as input for H3K27ac ? I applied the MACS14 to two inputs using the same parameters, one has 10 times cell numbers ( 110^6 ) , the generated bed could be used for downstream analysis, but the low cell number inputs had some issue with downstream analysis.
Thank you for any suggestion and comments. (simple links to relavent literature would be appreciated)