Hello,

I am using scanpy.external.pp.hashsolo to demultiplex my samples based on hashtags. However, I noticed that if I put the HTOs in different order I get slightly different results. Is that expected behavior?

sce.pp.hashsolo(sample_adata2,["Hash-tag1", "Hash-tag12", "Hash-tag13", "Hash-tag14", "Hash-tag2", "Hash-tag20", "Hash-tag3", "Hash-tag4"])
sce.pp.hashsolo(sample_adata2,["Hash-tag1", "Hash-tag2", "Hash-tag3", "Hash-tag4", "Hash-tag12", "Hash-tag13", "Hash-tag14", "Hash-tag20"])

I get between 5-20 more cells for each hashtag when I run it the first way.

Thank you