No p-value and adjusted p-value in adata.uns['cosg']

[chc1234567890]

Hi,

I am using the cosg package for differentially expressed genes identification in my research. However, I noticed that the output does not include p-values and adjusted p-values, which are crucial for subsequent analysis, e.g. pathway over-representation analysis. It seems that they haven't been implemented in the source code. Also, It would greatly enhance the usability of the package if multiple methods for p-value and adjusted p-value calculation can be added (like those in scanpy.tl.rank_genes_groups, e.g. t-test, wilcoxon for p-value and benjamini-hochberg for adjusted p-value).

Thank you for considering this feature request!

sc.logging.print_header()
# scanpy==1.10.3 anndata==0.11.1 umap==0.5.7 numpy==1.24.3 scipy==1.13.1 pandas==2.2.3 scikit-learn==1.5.2
# statsmodels==0.14.4 igraph==0.11.5 pynndescent==0.5.13

cosg.cosg(adata,
    key_added='cosg',
        mu=1,
        n_genes_user=2000,
               groupby=group_var)

adata.uns['cosg']['pvals'] # Not implemented yet

adata.uns['cosg']['pvals_adj'] # Not implemented yet