Downsampling ATAC-seq BAM files

Question

[Elise]

Hi all,

I have two ATAC datasets sequenced on two sequencers with different read depths. I would like to downsample the one with higher read depth to match that of the other dataset as we are observing batch effects following integration.

To my understanding, I need to shuffle the BAM files to equal depth, and then I'll need to convert the BAM files to a fragment file for the purposes of my analysis. Unfortunately, I have no experience dealing with this (only worked with RNA-seq before). If anyone is knowledgeable about the process and could elaborate on it, I would appreciate any insight.

Thank you in advance!

[ATpoint]

is this single-cell?

[Elise]

Single-nucleus as a part of the 10x multiomics protocol, yes.