...
#542
by
vasudev1717
- opened
This comment has been hidden
Thanks for your question! The "cls_and_gene" mode is for determining the impact of gene deletion on the cell embedding as a whole (cls) as well as the other genes in the cell (gene). If you are trying to perform perturbations comparing the impact of a gene on other genes, you do not need to provide cell_states_to_model and state_embs_dict, as this is not relevant to the gene to gene interaction. I would suggest trying without that if you are interested in quantifying gene shifts.
ctheodoris
changed discussion status to
closed
vasudev1717
changed discussion title from
Regarding aggregate_gene_shifts in in_silico_perturber_stats.py
to ...