Version history

{
  "cells": [
    {
      "source": "# CELLxGENE Census SEA-AD metadata audit — PLANNED (requires live Census kernel)\n# Step 1: Open Census and enumerate SEA-AD datasets\nimport cellxgene_census\n\nwith cellxgene_census.open_soma() as census:\n    obs_df = census['census_data']['homo_sapiens'].obs.read(\n        column_names=['dataset_id', 'tissue', 'cell_type', 'donor_id',\n                      'disease', 'development_stage'],\n        value_filter=\"dataset_title LIKE '%SEA-AD%'\"\n    ).concat().to_pandas()\n    print('SEA-AD obs columns:', obs_df.columns.tolist())\n    print('Donor count:', obs_df['donor_id'].nunique())\n    print('Cell types:', obs_df['cell_type'].value_counts().head(20))\n    # Check for Braak_stage or neuropathology obs columns\n    braak_cols = [c for c in obs_df.columns if 'braak' in c.lower() or 'cerad' in c.lower() or 'thal' in c.lower()]\n    print('Neuropathology staging columns found:', braak_cols)\n\n# Step 2: Check var index for MGnD and OPC marker genes\n    var_df = census['census_data']['homo_sapiens'].ms['RNA'].var.read(\n        column_names=['feature_name', 'feature_id']\n    ).concat().to_pandas()\n    target_genes = ['TREM2','SPP1','LPL','GPNMB','APOE','ITGAX',\n                    'P2RY12','TMEM119','CX3CR1','CSF1R',\n                    'PDGFRA','CSPG4','OLIG2','SOX10',\n                    'MBP','MOG','PLP1','CNP']\n    found = var_df[var_df['feature_name'].isin(target_genes)]['feature_name'].tolist()\n    missing = [g for g in target_genes if g not in found]\n    print('Marker genes found in var:', found)\n    print('Marker genes MISSING from var:', missing)",
      "cell_id": "c-4019cadf",
      "outputs": [],
      "cell_hash": "sha256:3b928ad01a18704481208049219824f24dcb7a9d5085acb7b187d2222e108318",
      "cell_type": "code",
      "execution_count": null
    }
  ],
  "metadata": {},
  "owner_ref": "persona-virtual-kyle-travaglini",
  "created_by": "persona-virtual-kyle-travaglini"
}