search

kaizhang / SnapATAC2

Single-cell epigenomics analysis tools
https://kzhang.org/SnapATAC2/
223 stars 25 forks source link

Question on peak calling procedure. #162

Closed UCDNJJ closed 11 months ago

UCDNJJ commented 1 year ago

Hi, looking to move from ArchR to SnapATAC2. I'm mostly interested to know how, if at all, SnapATAC2 differs from ArchR at the peak calling step. I tried figuring this out myself via your tutorials and code but wasn't clear on a few points:

Thanks!

kaizhang commented 1 year ago

Hi, looking to move from ArchR to SnapATAC2. I'm mostly interested to know how, if at all, SnapATAC2 differs from ArchR at the peak calling step. I tried figuring this out myself via your tutorials and code but wasn't clear on a few points:

  • Does SnapATAC2 call peak with MACS2 on pseudo-bulk measurements or single cell?

Currently yes!

  • How does SnapATAC2 handle replicates when calling / filtering peaks?

SnapATAC2 can call peaks in each replicate separately and then the peaks found in both replicates are retained.

  • Is it possible to recover the peak summit position from the MACS2 call within the anndata object?

Not possible now but I'm working on it.

I'm in the process of improving the peak calling procedure. Please feel free to propose new ideas and follow the development in this issue: #142

UCDNJJ commented 11 months ago

Really great work on improving the peak calling procedures! Thanks!