Statistical model implementation in MACS

[Moshe Olshansky]

Hello Tao,

Is there any possibility to get more detail about the actual implementation of the statistical model in MACS?

For example, when treatment and control pileups (normalised) are used, I understand that a (hard-coded) window of 200 bp is used. For this window, how are treatment and pileup compared (i.e. the actual statistic)?

And for bdgfiff, how is this done?

Is there any document describing this?

Thank you,

Moshe.