bonus checkpoint

src/talos/RunHailFiltering.py

@@ -43,7 +43,7 @@
 MAX_PARTITIONS = 10000
 
 
-def annotate_talos_clinvar(mt: hl.MatrixTable, clinvar: str) -> hl.MatrixTable:
+def annotate_clinvarbitration(mt: hl.MatrixTable, clinvar: str) -> hl.MatrixTable:
     """
     Don't allow these annotations to be missing
     - Talos has been co-developed with ClinvArbitration, a ClinVar re-summary effort
@@ -852,6 +852,14 @@ def main(
     mt = hl.read_matrix_table(mt_path)
     get_logger().info(f'Loaded annotated MT from {mt_path}, size: {mt.count_rows()}, partitions: {mt.n_partitions()}')
 
+    # lookups for required fields all delegated to the hail_audit file
+    if not (
+        fields_audit(mt=mt, base_fields=BASE_FIELDS_REQUIRED, nested_fields=FIELDS_REQUIRED)
+        and vep_audit(mt=mt, expected_fields=VEP_TX_FIELDS_REQUIRED)
+    ):
+        mt.describe()
+        raise KeyError('Fields were missing from the input Matrix')
+
     # repartition if required - local Hail with finite resources has struggled with some really high (~120k) partitions
     # this creates a local duplicate of the input data with far smaller partition counts, for less processing overhead
     if mt.n_partitions() > MAX_PARTITIONS:
@@ -861,14 +869,6 @@ def main(
             get_logger().info('Trying to write the result locally, might need more space on disk...')
             mt = generate_a_checkpoint(mt, f'{checkpoint}_reparitioned')
 
-    # lookups for required fields all delegated to the hail_audit file
-    if not (
-        fields_audit(mt=mt, base_fields=BASE_FIELDS_REQUIRED, nested_fields=FIELDS_REQUIRED)
-        and vep_audit(mt=mt, expected_fields=VEP_TX_FIELDS_REQUIRED)
-    ):
-        mt.describe()
-        raise KeyError('Fields were missing from the input Matrix')
-
     # subset to currently considered samples
     mt = subselect_mt_to_pedigree(mt, pedigree=pedigree)
 
@@ -878,15 +878,15 @@ def main(
     # remove any rows which have no genes of interest
     mt = remove_variants_outside_gene_roi(mt=mt, green_genes=green_expression)
 
+    if checkpoint:
+        mt = generate_a_checkpoint(mt, f'{checkpoint}_green_genes')
+
     # swap out the default clinvar annotations with private clinvar
-    mt = annotate_talos_clinvar(mt=mt, clinvar=clinvar)
+    mt = annotate_clinvarbitration(mt=mt, clinvar=clinvar)
 
     # remove common-in-gnomad variants (also includes ClinVar annotation)
     mt = filter_to_population_rare(mt=mt)
 
-    if checkpoint:
-        mt = generate_a_checkpoint(mt, f'{checkpoint}_data')
-
     # filter out quality failures
     mt = filter_on_quality_flags(mt=mt)
 
@@ -902,6 +902,9 @@ def main(
     # split each gene annotation onto separate rows, filter to green genes (PanelApp ROI)
     mt = split_rows_by_gene_and_filter_to_green(mt=mt, green_genes=green_expression)
 
+    if checkpoint:
+        mt = generate_a_checkpoint(mt, f'{checkpoint}_green_and_clean')
+
     # add Labels to the MT
     # current logic is to apply 1, 2, 3, and 5, then 4 (de novo)
     # for cat. 4, pre-filter the variants by tx-consequential or C5==1

test/test_hail_categories.py

@@ -14,7 +14,7 @@
     annotate_category_3,
     annotate_category_5,
     annotate_category_6,
-    annotate_talos_clinvar,
+    annotate_clinvarbitration,
     filter_to_categorised,
     filter_to_population_rare,
     green_from_panelapp,
@@ -306,7 +306,7 @@ def test_annotate_talos_clinvar(rating, stars, rows, regular, strong, tmp_path,
     table_path = str(tmp_path / 'anno.ht')
     table.write(table_path)
 
-    returned_table = annotate_talos_clinvar(make_a_mt, clinvar=table_path)
+    returned_table = annotate_clinvarbitration(make_a_mt, clinvar=table_path)
     assert returned_table.count_rows() == rows
     assert len([x for x in returned_table.info.clinvar_talos.collect() if x == 1]) == regular
     assert len([x for x in returned_table.info.clinvar_talos_strong.collect() if x == 1]) == strong