Cell cycle regression

[Robinovic]

Hi,
First of all, thanks for the very useful tool and the documentation very clear.
I would like to regress out effect of phase's heterogeneity among cells in my dataset. In #173, you proposed to use scanpy.pp.regress_out function to remove effect of cell cycle. According to the documentation, this function return the corrected count matrix belonging only to adata.X. As velocity is computed based on layers 'spliced' and 'unspliced', regressing out cell cycle in this manner is not very useful for me.
I am looking for a method to regress cell cycle effect on adata.X and other layers of Anndata object. I could simply delete the cell cycle genes from selected genes, but I will probably regressed the variable G2M.score-S.score to keep heterogeneity between G1 and proliferative cells.

Edit: I was just looking for something too complicated, simply regressing cell cycle effect or any effect you want to remove with scanpy is enough, no need to regress on spliced and unspliced matrix.
sc.pp.regress_out(adata, keys=['S_score','G2M_score'] )

Once again, thanks for the great tool.

Robin