When running finalize-dataloader on a dataset, zero features get mapped to the reference IDs if feature_id_var_key is set in the yaml and the ensmbl IDs in the adata contain gene versions (eg. ENSG00000131584.19). The version used to be filtered out automatically before and the docstrings/function attributes for the mapping still suggest that this is done. It's however not.
When running finalize-dataloader on a dataset, zero features get mapped to the reference IDs if
feature_id_var_keyis set in the yaml and the ensmbl IDs in the adata contain gene versions (eg.ENSG00000131584.19). The version used to be filtered out automatically before and the docstrings/function attributes for the mapping still suggest that this is done. It's however not.