Cell-Cycle Variation in 10x Multiomics

Question

[georgii.vdovin]

Hello,

I have 10x ATAC+RNA single nuclei data from 10x, for T cells. There is a distinct cluster of proliferating T cells, and for cell subtype annotation I could do a S/G2M Phase regression on the RNA part. However, when running FindClusters(pbmc, graph.name = "wsnn", algorithm = 3) with Seurat, the ATAC features are also used, so the cell-cycle variation is preserved in the clusters coming from the ATAC part.

My question is conceptual: what is the best practice of dealing with Cell Cycle variation in multiomics? Simply preserving it? Do people ever attempt to regress it out from both RNA + ATAC?

Thanks in advance.