Counts in layers is zero after ad.concat

Hi,

I filter each dataset before ad.concat:

sc.pp.filter_genes(adata, min_cells = 3)
sc.pp.filter_cells(adata, min_genes = 200)
sc.pp.filter_cells(adata, min_counts = 200)

I also tried

sc.pp.filter_cells(adata, min_counts = 200)
sc.pp.filter_genes(adata, min_counts=3)
sc.pp.filter_genes(adata, min_cells = 3)

Then

ad.experimental.concat_on_disk(in_files = infiles, merge = 'same', out_file = '10x_all.h5ad')

But some of the sub datasets return zero counts in layers. Why do I get zero counts when I’ve already filtered the sub datasets to have >= 200 genes? Any help would be much appreciated.

I figured out that using ad.experimental.concat_on_disk caused the zero counts. Reverting to ad.concat solved the problem.

Could you open an issue on anndata for this? It seems like a bug in concat_on_disk