Sorry if these questions may seem a bit dumb, I am very new to CHiP-seq and bioinformatics in general. I am trying to use DiffBind to find differential binding between groups.
What if a peak is present for a gene in one group but not the other? How would the program take that into account for calculating differential binding?
I am also very confused regarding occupancy analysis and differential binding analysis. What exactly differentiates the two in a biological context?
The DiffBind vignette attempts to explain how this is handled in step-by-step detail, that is another good place to start.
The direct answer to your question is that in a simple occupancy analysis, only binding sites identified in one sample group and not the other can be identified as somehow "differential". In a quantitative occupancy analysis, the read counts are determined for every potential binding site across all samples regardless of whether a peak was identified in that sample. In that way, intervals identified as a peak in one sample but not another, but where the (normalized) read counts are very similar, may turn out have a low confidence of being differentially bound, while a peak identified across all sample groups, but where the read density differs systematically between sample groups, may be identified with high confidence as being differential.
I'm very new myself and there's plenty of people who can answer your question better than I can, but here is a resource I found extremely helpful:
https://github.com/hbctraining/Intro-to-ChIPseq
I think they just rolled out with a newer version too which might be linked there.
I've moved this to a comment as it does not directly answer OP's question.