Questions on batch correction with exp matrix (generation of .table file)

[mapo121]

This is a question about the .table file. I understand that LIGER integrates the scATAC and scRNA-seq. What exactly does this .table contain, and how can I generate it from other pipelines (Seruat, scanpy). Do I need to also use an ATAC pipline like Signac?

I've been using Seurat to process my sample (the scRNA-seq part) -how does scMTNI handle batch correction? Supposing if I have an integrated scRNA-seq sample, how does that effect the downstream analysis?

Should we avoid integrating samples for scMTNI (scRNA-seq wise)? Also, what assay should we use when calculating the expression matrix for each cluster? Should we use just the counts? Should we apply some normalization?

[sroyyors]

scMTNI is not for integrating scATAC-seq matrices from different samples or treatments. It expects that these are already done before you give it to scMTNI using tools like LIGER or Signac or whatever you prefer. For scRNA-seq the .table files are the expression matrices per cluster. You could get these from Seurat clusters.