obs_level_CT_rank
Function
crisgi_obj.obs_level_CT_rank(gene_sets, prefix='test', min_size=5)
Performs observation-level cell type (CT) ranking using gene set variation analysis (GSVA) on the provided gene sets. This method is designed as a member function of the crisgi_obj class. It computes enrichment scores for each observation (cell/sample), ranks them, and saves the results to a CSV file. If group information is available, it annotates the results accordingly.
Parameters
| Name | Type | Description |
|---|---|---|
| gene_sets | object | Dictionary or compatible object containing gene sets for GSVA analysis. |
| prefix | str | Prefix for output directory and files. Default is 'test'. |
| min_size | int | Minimum number of genes required in each gene set. Default is 5. |
Return type
pandas.DataFrame
Returns
A DataFrame with GSVA enrichment scores for each observation, optionally annotated with group information and sorted by enrichment score (ES). The results are also exported as a CSV file in the specified output directory.
Attributes Set
self.gp_es: Stores the GSVA result object for downstream analysis.
Example
# Assuming `crisgi_obj` is an instance with `edata`, `out_dir`, and optionally `groupby` attributes
gene_sets = {'Pathway1': ['GeneA', 'GeneB'], 'Pathway2': ['GeneC', 'GeneD']}
result_df = crisgi_obj.obs_level_CT_rank(gene_sets=gene_sets, prefix='experiment1', min_size=10)
# The results are saved to './<out_dir>/experiment1/prerank_gsva_interaction.csv'
# The DataFrame `result_df` contains the ranked enrichment scores for each observation.