I am learning the DiffBind package, and I have some questions not covered by the vignette.
1) Is the peak score relevant to affinity analysis, and in what way if so?
2) In what case (generally) one should use an occupancy analysis and not an affinity analysis?
3) The idea of greylists is to exclude area where a disproportionate degree of signal is shown at the control. Should not using the control sample at the peak caller stage be sufficient? If there is a very strong signal in the control, then the peak caller should not declare a peak in that area.
4) What are the supplied control files used for (in affinity analysis)? I understand that they will be used if one performs background normalization; is there another use for them?