Get unnormal spliced/unspliced proportion

[HuangDDU]

...
I use scvelo on my data and get unnormal spliced/unspliced proportion. Can I continue scVelo proportion on my data.

# paste your code here, if applicable
scv.pl.proportions(adata,save="proportion.pdf")


The proportion of spliced and unspliced mRNA is 2:8, which is just oppsite to the wanted condition of 8:2

# paste the error output here, if applicable
The proportion of spliced and unspliced mRNA is 2:8, which  is just oppsite to the wanted condition of 8:2.

Versions

# paste the ouput of scv.logging.print_versions() here
scvelo==0.2.4  scanpy==1.9.1  anndata==0.8.0  loompy==3.0.7  numpy==1.21.5  scipy==1.8.0  matplotlib==3.5.2  sklearn==1.0.2  pandas==1.4.2 

[WeilerP]

@HuangDDU, yes, in theory, you can continue. I'd be cautious, though. A Couple of thoughts:

1. scv.pl.proportions(adata) is intended to be used with raw counts. Otherwise, this results in unexpected behavior.
2. To understand if RNA velocity can be inferred for your data, you need to have a look at the phase portraits. See e.g. about gene for velocity #216 or Velocity direction reversed from biological direction #462.