scib.metrics.kBET

scib.metrics.kBET(adata, batch_key, label_key, scaled=True, embed='X_pca', type_=None, return_df=False, verbose=False)

kBET score

Compute the average of k-nearest neighbour batch effect test (kBET) score per label.

Parameters

  • adata – anndata object to compute kBET on

  • batch_key – name of batch column in adata.obs

  • label_key – name of cell identity labels column in adata.obs

  • scaled – whether to scale between 0 and 1 with 0 meaning low batch mixing and 1 meaning optimal batch mixing if scaled=False, 0 means optimal batch mixing and 1 means low batch mixing

Returns

kBET score (average of kBET per label) based on observed rejection rate. If return_df=True, also return a pd.DataFrame with kBET observed rejection rate per cluster