Running scCODA with confounding covariate

Help pertpy
1

Hi!

I’m running scCODA on an integrated AnnData object with integrated multiple datasets, and I have a few questions:

1.My main variable of interest is condition (i.e.,adata.obs["condition"]). Should I include dataset — which is equivalent to a batch variable and is also stored in adata.obs — as a confounding covariate?

2.If so, should I specify both variables in the covariate_obs argument of the load() function, like this:

covariate_obs=["condition", "dataset"]

and then set the formula argument in prepare() as follows?

formula="C(condition, Treatment('Disease')) + dataset"

Thanks in advance for your help!