scglue.models.dx.integration_consistency(model, adatas, graph, edge_weight='weight', edge_sign='sign', count_layers=None)[source]

Integration consistency score, defined as the consistency between aligned-space meta-cell correlation and the prior graph

  • model (SCGLUEModel) – Integration model to be evaluated

  • adatas (Mapping[str, AnnData]) – Datasets (indexed by domain name)

  • graph (Graph) – Prior graph

  • edge_weight (str) – Key of edge attribute for edge weight

  • edge_sign (str) – Key of edge attribute for edge sign

  • count_layers (Optional[Mapping[str, str]]) – Dataset layers that contain raw counts (indexed by domain name). If not provided, it is assumed that .X is the raw counts.

Return type



consistency_df – Consistency score at different numbers of meta cells