Import annotation transfer onto single cells

[dosumis]

Support import of annotation transfer from MapMyCells JSON or from CSV.

CSV must have cell ID column, transferred label, optionally heirarchy level of transferred label. Optionally confidence score.

• Generates a new labelset - unranked. Import should support specification of general metadata about annotation transfer, including source taxonomy and method used.
• Cell set membership defined by links to cell_IDs
• Schema extension on CAS support recording confidence - LInk TBA

[dosumis]

DRAFT

MapMyCells JSON from hierarchical mapping

  "results": [
    {
      "CCN20230722_CLAS": {
        "assignment": "CS20230722_CLAS_01",
        "bootstrapping_probability": 1.0,
        "avg_correlation": 0.5811543426123422,
        "runner_up_assignment": [],
        "runner_up_correlation": [],
        "runner_up_probability": []
      },
      "CCN20230722_SUBC": {
        "assignment": "CS20230722_SUBC_022",
        "bootstrapping_probability": 1.0,
        "avg_correlation": 0.497930470296775,
        "runner_up_assignment": [],
        "runner_up_correlation": [],
        "runner_up_probability": []
      },
      "CCN20230722_CLUS": {
        "assignment": "CS20230722_CLUS_0364",
        "bootstrapping_probability": 1.0,
        "avg_correlation": 0.5401797101798183,
        "runner_up_assignment": [],
        "runner_up_correlation": [],
        "runner_up_probability": []
      },
      "cell_id": "20171204_sample_2"
    },
...
  "marker_genes": {
    "CCN20230722_CLAS/CS20230722_CLAS_01": [
      "ENSMUSG00000051951",
      "ENSMUSG00000002459",
      "ENSMUSG00000025905",
      "ENSMUSG00000033740",
      "ENSMUSG00000042501",
      ...
 "taxonomy_tree": {
    "metadata": {
      "factory": "from_data_release",
      "timestamp": "2023-08-21-13-13-48",
      "params": {
        "cell_metadata_path": null,
        "cluster_annotation_path": "/allen/programs/celltypes/workgroups/rnaseqanalysis/lydian/ABC_handoff/metadata/WMB-taxonomy/20230830/cluster_annotation_term.csv",
        "cluster_membership_path": "/allen/programs/celltypes/workgroups/rnaseqanalysis/lydian/ABC_handoff/metadata/WMB-taxonomy/20230830/cluster_to_cluster_annotation_membership.csv",
        "hierarchy": [
          "CCN20230722_CLAS",
          "CCN20230722_SUBC",
          "CCN20230722_SUPT",
          "CCN20230722_CLUS"
        ]
      }
    },
...

In this case we need to create 3 labelsets, one for each hierarchy level

CAS labelset metadata field:

Each labelset