stereo.core.StPipeline.find_marker_genes#

StPipeline.find_marker_genes(cluster_res_key, method='t_test', case_groups='all', control_groups='rest', corr_method='benjamini-hochberg', use_raw=True, use_highly_genes=True, hvg_res_key='highly_variable_genes', res_key='marker_genes', output=None, sort_by='scores', n_genes='all', ascending=False, n_jobs=4)[source]#

A tool to find maker genes. For each group, find statistical test different genes between one group and the rest groups using t_test or wilcoxon_test.

Parameters:

  • cluster_res_key – the key of clustering to get corresponding result from self.result.

  • method (Literal['t_test', 'wilcoxon_test']) – choose method for statistics.

  • case_groups (Union[str, ndarray, list]) – case group, default all clusters.

  • control_groups (Union[str, ndarray, list]) – control group, default the rest of groups.

  • corr_method (str) – correlation method.

  • use_raw (bool) – whether to use raw express matrix for analysis, default True.

  • use_highly_genes (bool) – whether to use only the expression of hypervariable genes as input, default True.

  • hvg_res_key (Optional[str]) – the key of highly variable genes to get corresponding result.

  • res_key (str) – the key for storing result of marker genes.

  • output (Optional[str]) – the path to output file .csv. If None, do not generate output file.

  • sort_by – default to ‘scores’, the result will sort by the key, other options ‘log2fc’.

  • n_genes (Union[str, int]) – default to 0, means will auto calculate n_genes by N = 10000/K². K is cluster number, and N is larger or equal to 1, less or equal to 50.

  • ascending (bool) – default to False.

  • n_jobs (int) – the number of parallel jobs to run. default to 4.

Returns:

The result of marker genes is stored in self.result where the key is 'marker_genes'.