I am trying to use Scrublet as a preprocessing step in my pipeline. I use a filtered anndata object with raw UMI counts.
If I pass normalize_variance = True while also passing copy = False when running scanpy.external.pp.scrublet(), does this mean that my UMI counts of the anndata object will be normalized and replace these raw UMI counts? Or does normalize_variance leave these UMI counts unaffected?
It is important that the UMI counts remain raw.
Hello Scrublet team,
I am trying to use Scrublet as a preprocessing step in my pipeline. I use a filtered anndata object with raw UMI counts. If I pass
normalize_variance = Truewhile also passingcopy = Falsewhen runningscanpy.external.pp.scrublet(), does this mean that my UMI counts of the anndata object will be normalized and replace these raw UMI counts? Or doesnormalize_varianceleave these UMI counts unaffected? It is important that the UMI counts remain raw.Thank you in advance,
Kind regards,
Julia