awb WIP (#444)

.gitignore

@@ -52,3 +52,7 @@ _templates
 
 # script output
 /files/out
+
+#python
+venv/
+

awb/list.py

@@ -0,0 +1,125 @@
+import anndata as ad
+import pandas as pd
+
+# Load your AnnData file
+adata = ad.read_h5ad("/Users/dave/Downloads/skin.h5ad")
+
+# Columns categorized as CELLxGENE Required, HCA Required, and HCA Recommended
+cellxgene_required = [
+    "assay_ontology_term_id",
+    "cell_type_ontology_term_id",
+    "development_stage_ontology_term_id",
+    "disease_ontology_term_id",
+    "donor_id",
+    "is_primary_data",
+    "organism_ontology_term_id",
+    "self_reported_ethnicity_ontology_term_id",
+    "sex_ontology_term_id",
+    "suspension_type",
+    "tissue_type",
+    "tissue_ontology_term_id",
+]
+
+hca_required = [
+    "alignment_software",
+    "cell_enrichment",
+    "gene_annotation_version",
+    "institute",
+    "library_id",
+    "library_preparation_batch",
+    "library_sequencing_run",
+    "manner_of_death",
+    "reference_genome",
+    "sampled_site_condition",
+    "sample_collection_method",
+    "sample_id",
+    "sample_preservation_method",
+    "sample_source",
+    "sequenced_fragment",
+]
+
+hca_recommended = [
+    "author_batch_notes",
+    "author_cell_type",
+    "cell_number_loaded",
+    "cell_viability_percentage",
+    "intron_inclusion",
+    "library_id_repository",
+    "protocol_url",
+    "sample_collection_relative_time_point",
+    "sequencing_platform",
+    "sample_collection_site",
+    "sample_collection_year",
+    "tissue_free_text",
+]
+
+
+# Function to determine distinct values and null percentages for the columns in obs
+def count_distinct_values(adata, columns):
+    distinct_counts = {}
+    for col in columns:
+        if col in adata.obs.columns:
+            distinct_counts[col] = {
+                "distinct_count": adata.obs[col].nunique(),
+                "first_values": adata.obs[col].unique()[:5].tolist(),
+                "null_percentage": adata.obs[col].isnull().mean() * 100,
+            }
+        else:
+            distinct_counts[col] = {
+                "distinct_count": "Column not found",
+                "first_values": "Column not found",
+                "null_percentage": "Column not found",
+            }
+    return distinct_counts
+
+
+# Get distinct value counts for CELLxGENE required, HCA required, and recommended columns
+cellxgene_required_counts = count_distinct_values(adata, cellxgene_required)
+required_counts = count_distinct_values(adata, hca_required)
+recommended_counts = count_distinct_values(adata, hca_recommended)
+
+# Convert results to DataFrame for better display
+cellxgene_required_df = pd.DataFrame(
+    [
+        (col, data["distinct_count"], data["first_values"], data["null_percentage"])
+        for col, data in cellxgene_required_counts.items()
+    ],
+    columns=[
+        "CELLxGENE Required Columns",
+        "Distinct Values",
+        "First 5 Distinct Values",
+        "Null Percentage",
+    ],
+)
+required_df = pd.DataFrame(
+    [
+        (col, data["distinct_count"], data["first_values"], data["null_percentage"])
+        for col, data in required_counts.items()
+    ],
+    columns=[
+        "HCA Required Columns",
+        "Distinct Values",
+        "First 5 Distinct Values",
+        "Null Percentage",
+    ],
+)
+recommended_df = pd.DataFrame(
+    [
+        (col, data["distinct_count"], data["first_values"], data["null_percentage"])
+        for col, data in recommended_counts.items()
+    ],
+    columns=[
+        "HCA Recommended Columns",
+        "Distinct Values",
+        "First 5 Distinct Values",
+        "Null Percentage",
+    ],
+)
+
+# Display the results
+print("CELLxGENE Required Columns Distinct Values:\n")
+print(cellxgene_required_df)
+print("\nHCA Required Columns Distinct Values:\n")
+print(required_df)
+print("\nHCA Recommended Columns Distinct Values:\n")
+print(recommended_df)

awb/requirements.txt

@@ -0,0 +1,4 @@
+anndata>=0.8,<0.12
+numpy<2
+pandas>2,<3
+