morris-lab / CellOracle

This is the alpha version of the CellOracle package
Other
309 stars 56 forks source link

How to process multiple Bulk ATAC-seq datasets? #135

Open Crorazon opened 1 year ago

Crorazon commented 1 year ago

Hi @KenjiKamimoto-wustl122,

Thank you for sharing this useful package with us. I have already learned how to process a Bulk ATAC-seq data. However, I have 32 bulk ATAC-seq datasets here, and I am wondering whether I should merge them for processing.

yangjun9095 commented 1 year ago

Hi! I'm not a CellOracle developer, but wanted to leave my comment here.

Since ATAC-seq is used to compute cis-regulatory elements for each gene (enhancers), I think there can be a couple of options here. 1) process each bulk ATAC-seq data independently, computing peaks here, and finding the joint peak profile by "merge" using Signac package. 2) integrate multiple bulk ATAC-seq datasets (I don't know exactly how), then compute the peak profiles from the integrated datasets.

It also depends on the heterogeneity of your 32 bulk ATAC-seq datasets. For example, whether they are biological replicates, or from different conditions. If they come from different conditions, I would probably integrate only among the same conditions.

Best, Yang-Joon

Crorazon commented 1 year ago

@yangjun9095

Thank you again for your suggestions. I think the first approach of processing each dataset independently has good feasibility. However, I have one question - can the Signac package handle bulk ATAC-seq data? From my understanding, it seems suited more for scRNA-seq data.

Would you please let me know if Signac is indeed capable of merging peaks from multiple bulk ATAC-seq samples? If not, do you have any other recommendations on software tools I could use for this purpose?