Merge pull request #4417 from broadinstitute/hail-search-sex-meatdata

Use boolean for sending sex metadata to hail search
broadinstitute · Oct 15, 2024 · 2c9553b · 2c9553b
2 parents 2f3473d + 5de0094
commit 2c9553b
Show file tree

Hide file tree

Showing 4 changed files with 18 additions and 19 deletions.
diff --git a/hail_search/constants.py b/hail_search/constants.py
@@ -7,7 +7,6 @@
 AFFECTED_ID = 0
 UNAFFECTED_ID = 1
 UNKNOWN_AFFECTED_ID = 2
-MALE = 'M'
 AFFECTED_ID_MAP = {AFFECTED: AFFECTED_ID, UNAFFECTED: UNAFFECTED_ID, UNKNOWN_AFFECTED: UNKNOWN_AFFECTED_ID}
 
 GROUPED_VARIANTS_FIELD = 'variants'

diff --git a/hail_search/queries/base.py b/hail_search/queries/base.py
@@ -6,7 +6,7 @@
 
 from hail_search.constants import AFFECTED_ID, ALT_ALT, ANNOTATION_OVERRIDE_FIELDS, ANY_AFFECTED, COMP_HET_ALT, \
     COMPOUND_HET, GENOME_VERSION_GRCh38, GROUPED_VARIANTS_FIELD, ALLOWED_TRANSCRIPTS, ALLOWED_SECONDARY_TRANSCRIPTS,  HAS_ANNOTATION_OVERRIDE, \
-    HAS_ALT, HAS_REF,INHERITANCE_FILTERS, PATH_FREQ_OVERRIDE_CUTOFF, MALE, RECESSIVE, REF_ALT, REF_REF, MAX_LOAD_INTERVALS, \
+    HAS_ALT, HAS_REF,INHERITANCE_FILTERS, PATH_FREQ_OVERRIDE_CUTOFF, RECESSIVE, REF_ALT, REF_REF, MAX_LOAD_INTERVALS, \
     UNAFFECTED_ID, X_LINKED_RECESSIVE, XPOS, OMIM_SORT, FAMILY_GUID_FIELD, GENOTYPES_FIELD, AFFECTED_ID_MAP
 
 HAIL_SEARCH_DATA_DIR = os.environ.get('HAIL_SEARCH_DATA_DIR', '/seqr/seqr-hail-search-data')
@@ -563,7 +563,7 @@ def _sample_entry_data(cls, sample):
             sampleId=sample['sample_id'],
             individualGuid=sample['individual_guid'],
             affected_id=AFFECTED_ID_MAP.get(sample['affected']),
-            is_male='sex' in sample and sample['sex'] == MALE,
+            is_male=sample.get('is_male', False),
         )
 
     @classmethod

diff --git a/hail_search/test_utils.py b/hail_search/test_utils.py
@@ -6,28 +6,28 @@
     'project_guid': 'R0001_1kg', 'affected': 'A', 'sample_type': 'WES',
 }
 FAMILY_2_VARIANT_SAMPLE_DATA_WITH_SEX = {'SNV_INDEL': [
-    {'sample_id': 'HG00731', 'individual_guid': 'I000004_hg00731', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'A', 'sample_type': 'WES', 'sex': 'F'},
-    {'sample_id': 'HG00732', 'individual_guid': 'I000005_hg00732', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'sex': 'M'},
-    {'sample_id': 'HG00733', 'individual_guid': 'I000006_hg00733', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'sex': 'F'},
+    {'sample_id': 'HG00731', 'individual_guid': 'I000004_hg00731', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'A', 'sample_type': 'WES', 'is_male': False},
+    {'sample_id': 'HG00732', 'individual_guid': 'I000005_hg00732', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'is_male': True},
+    {'sample_id': 'HG00733', 'individual_guid': 'I000006_hg00733', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'is_male': False},
 ]}
 FAMILY_2_VARIANT_SAMPLE_DATA = deepcopy(FAMILY_2_VARIANT_SAMPLE_DATA_WITH_SEX)
 for s in FAMILY_2_VARIANT_SAMPLE_DATA['SNV_INDEL']:
-    s.pop('sex')
+    s.pop('is_male')
 
 EXPECTED_SAMPLE_DATA_WITH_SEX = {
     'SV_WES': [
-        {'sample_id': 'HG00731', 'individual_guid': 'I000004_hg00731', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'A', 'sample_type': 'WES', 'sex': 'F'},
-        {'sample_id': 'HG00732', 'individual_guid': 'I000005_hg00732', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'sex': 'M'},
-        {'sample_id': 'HG00733', 'individual_guid': 'I000006_hg00733', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'sex': 'F'}
+        {'sample_id': 'HG00731', 'individual_guid': 'I000004_hg00731', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'A', 'sample_type': 'WES', 'is_male': False},
+        {'sample_id': 'HG00732', 'individual_guid': 'I000005_hg00732', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'is_male': True},
+        {'sample_id': 'HG00733', 'individual_guid': 'I000006_hg00733', 'family_guid': 'F000002_2', 'project_guid': 'R0001_1kg', 'affected': 'N', 'sample_type': 'WES', 'is_male': False}
     ],
 }
 EXPECTED_SAMPLE_DATA_WITH_SEX.update(FAMILY_2_VARIANT_SAMPLE_DATA_WITH_SEX)
-EXPECTED_SAMPLE_DATA_WITH_SEX['SNV_INDEL'].append({'sex': 'M', **FAMILY_3_SAMPLE})
+EXPECTED_SAMPLE_DATA_WITH_SEX['SNV_INDEL'].append({'is_male': True, **FAMILY_3_SAMPLE})
 
 EXPECTED_SAMPLE_DATA = deepcopy(EXPECTED_SAMPLE_DATA_WITH_SEX)
 for samples in EXPECTED_SAMPLE_DATA.values():
     for s in samples:
-        s.pop('sex')
+        s.pop('is_male')
 
 CUSTOM_AFFECTED_SAMPLE_DATA = {'SNV_INDEL': deepcopy(EXPECTED_SAMPLE_DATA['SNV_INDEL'])}
 CUSTOM_AFFECTED_SAMPLE_DATA['SNV_INDEL'][0]['affected'] = 'N'
@@ -79,14 +79,14 @@
 MULTI_PROJECT_MISSING_SAMPLE_DATA = deepcopy(FAMILY_2_MISSING_SAMPLE_DATA)
 MULTI_PROJECT_MISSING_SAMPLE_DATA['SNV_INDEL'].append(FAMILY_11_SAMPLE_WES)
 
-SV_WGS_SAMPLE_DATA_WITH_SEX = {'SV_WGS': [{'sex': 'M', **FAMILY_11_SAMPLE_WES, 'sample_type': 'WGS'}, {
-    'sample_id': 'NA20884', 'individual_guid': 'I000025_na20884', 'family_guid': 'F000011_11', 'project_guid': 'R0003_test', 'affected': 'N', 'sample_type': 'WGS', 'sex': 'M',
+SV_WGS_SAMPLE_DATA_WITH_SEX = {'SV_WGS': [{'is_male': True, **FAMILY_11_SAMPLE_WES, 'sample_type': 'WGS'}, {
+    'sample_id': 'NA20884', 'individual_guid': 'I000025_na20884', 'family_guid': 'F000011_11', 'project_guid': 'R0003_test', 'affected': 'N', 'sample_type': 'WGS', 'is_male': True,
 }, {
-    'sample_id': 'NA20883', 'individual_guid': 'I000035_na20883', 'family_guid': 'F000011_11', 'project_guid': 'R0003_test', 'affected': 'N', 'sample_type': 'WGS', 'sex': 'F',
+    'sample_id': 'NA20883', 'individual_guid': 'I000035_na20883', 'family_guid': 'F000011_11', 'project_guid': 'R0003_test', 'affected': 'N', 'sample_type': 'WGS', 'is_male': False,
 }]}
 SV_WGS_SAMPLE_DATA = deepcopy(SV_WGS_SAMPLE_DATA_WITH_SEX)
 for s in SV_WGS_SAMPLE_DATA['SV_WGS']:
-    s.pop('sex')
+    s.pop('is_male')
 
 SV_WES_SAMPLE_DATA = {'SV_WES': EXPECTED_SAMPLE_DATA['SV_WES'] + [FAMILY_3_SAMPLE]}
 

diff --git a/seqr/utils/search/hail_search_utils.py b/seqr/utils/search/hail_search_utils.py
@@ -1,11 +1,11 @@
 from collections import defaultdict
 
-from django.db.models import F, Min, Count
+from django.db.models import F, Min, Count, Case, When
 from urllib3.connectionpool import connection_from_url
 
 import requests
 from reference_data.models import Omim, GeneConstraint, GENOME_VERSION_LOOKUP
-from seqr.models import Sample, PhenotypePrioritization
+from seqr.models import Sample, PhenotypePrioritization, Individual
 from seqr.utils.search.constants import PRIORITIZED_GENE_SORT, X_LINKED_RECESSIVE
 from seqr.utils.xpos_utils import MIN_POS, MAX_POS
 from settings import HAIL_BACKEND_SERVICE_HOSTNAME, HAIL_BACKEND_SERVICE_PORT
@@ -141,7 +141,7 @@ def _get_sample_data(samples, inheritance_filter=None, inheritance_mode=None, **
         affected=F('individual__affected'),
     )
     if inheritance_mode == X_LINKED_RECESSIVE:
-        sample_values['sex'] = F('individual__sex')
+        sample_values['is_male'] = Case(When(individual__sex=Individual.SEX_MALE, then=True), default=False)
     sample_data = samples.order_by('guid').values('individual__individual_id', 'dataset_type', 'sample_type', **sample_values)
 
     custom_affected = (inheritance_filter or {}).pop('affected', None)