apply more stringent filtering

openproblems-bio · Sep 25, 2024 · f9cb326 · f9cb326
1 parent 8bd0a33
commit f9cb326
Showing 1 changed file with 3 additions and 8 deletions.
diff --git a/src/datasets/loaders/allen_brain_cell_atlas/script.py b/src/datasets/loaders/allen_brain_cell_atlas/script.py
@@ -65,6 +65,9 @@
     adata = ad.read_h5ad(
         TMP_DIR / f"expression_matrices/WMB-10Xv2/{VERSION}/WMB-10Xv2-{region}-raw.h5ad"
     )
+    sc.pp.filter_cells(adata, min_genes=5)
+    sc.pp.filter_cells(adata, min_counts=50)
+
     adata = adata[adata.obs_names.isin(obs.index)]
     adata.obs["region"] = region
     counts = adata.X
@@ -82,14 +85,6 @@
 adata = ad.concat(adatas, merge="first")
 del adatas
 
-print("Filtering data", flush=True)
-num_cells_per_gene = adata.layers["counts"].getnnz(axis=0)
-num_genes_per_cell = adata.layers["counts"].getnnz(axis=1)
-adata = adata[
-    num_genes_per_cell >= 1,
-    num_cells_per_gene >= 1
-].copy()
-
 print("Processing .obs")
 adata.obs = obs.loc[adata.obs.index]