Performance optimisation request [PERFORMANCE]

[TingTingShao]

Dear,

I only have snATAC-seq for now, and would like to try pycistopic to run.

Original files: Fragments.tsv.gz
number of files/sample: 295
number of cells per file/sample: around 2000-3000 after QC
Number of features before excluding the blacklist: 6062095 regions (divided chromsome in 500bp bin size)

The files have been preprocess in snapATAC2 and want to try more features provided in cistopic and SCENICplus.

I think the tutorial Using pycisTopic on simulated melanoma data is more suitful so far, but the count matrix is too large. Do you have any suggestions on how to process large count matrix or given without snRNA-seq data, which tutorial would be more suitable for me to get started?

Looking forward to your reply!
thanks!
TingTing

[SeppeDeWinter]

Hi @TingTingShao

I converted your issue to a discussion, because you are asking for advice (rather than describing a particular issue).
Sounds like an exciting project you have going on there.

At which particular step in the tutorial are you getting stuck?

All the best,

Seppe