I want to load a CITE-seq anndata file and as per the CITE-seq reference mapping tutorial I want to use the argument mask_protein_batches=5
.
However in the tutorial the anndata is loaded from the built in datasets (as below), and I’d like to load it from disk (it is actually the same data but I wrangled it to be a subset of genes with properly mapped gene ids).
I could not figure out how to do this for a dataset from disk, and searching in the repo this argument seems to only exist for loading built in datasets?
https://github.com/scverse/scvi-tools/search?q=mask_protein_batches
adata = scvi.data.pbmc_seurat_v4_cite_seq(
run_setup_anndata=False,
mask_protein_batches=5,
)