nucleosome positioning question

[juechenyang]

Hi Alicia,

I am trying to use MNase data to test nucleoatac. I have a question regarding to the prediction for nucleosome. So I found that many of the high resolution calls from "base.nucpos.bed" are highly aligned with the cross-correlation raw signal(base.nucleoatac_signal.bedgraph) rather than the smoothed signal(base.nucleoatac_signal.smooth.bedgraph).

I think, in the paper, it says that nucleosome position calls are based on a greedy algorithm finding local maxima. I just don't which signal you applied this algorithm on? and if not on smoothed signal, why you wanna do this transfer from raw to smooth?

I attache a image for your reference. Thank you for your time