search

martinjzhang / scDRS

Single-cell disease relevance score (scDRS)
https://martinjzhang.github.io/scDRS/
MIT License
114 stars 16 forks source link

About the H5AD input - raw counts or normalized+transformed? #103

Closed hopehan12 closed 2 months ago

hopehan12 commented 2 months ago

Hi,

Thanks for the great work! I have a question about the H5AD file for scDRS. I already processed my single-cell data in Seurat so I extracted RNA assay counts to h5ad file (raw counts). But I read the paper found out it needs to be normalized and transformed. Now I am trying to run with normalized and transformed h5ad but it gives me warning when I am running the compute-score function 'Loading data: WARNING: adata.X seems to be already log-transformed.'

I am just wondering which types of H5AD file should I provide as input? Raw counts or normalized+transformed? Please just let me know! Many thanks.

martinjzhang commented 2 months ago

You can just supply the raw count data and specify the flag --flag-raw-count True

hopehan12 commented 2 months ago

Thank you Martin! Really helpful :)