Merge pull request #229 from Clinical-Genomics/refactor_predict_sex

Refactor report's sex rows and coverage.get_samples_predicted_sex to use d4tools instead of pyd4

CHANGELOG.md

@@ -6,6 +6,7 @@
 - Speed up response by `coverage.d4_intervals_coverage` by replacing pyd4 lib with direct calls d4tools and multiprocessing
 - Removed 2 redundant functions in `meta.handle.bed.py`
 - `coverage.d4_interval_coverage` is using direct calls to d4tools to retrieve stats over an entire chromosome or a genomic interval
+- Reformat report sample' sex rows and coverage.get_samples_predicted_sex endpoint to use d4tools and not pyd4 for evaluating sample sex
 ### Fixed
 -  `coverage.d4_interval_coverage` endpoint crashing trying to computer coverage completeness over an entire chromosome
 

src/chanjo2/endpoints/coverage.py

@@ -14,14 +14,12 @@
 from chanjo2.dbutil import get_session
 from chanjo2.meta.handle_bed import bed_file_interval_id_coords
 from chanjo2.meta.handle_d4 import (
-    get_d4_file,
     get_d4tools_chromosome_mean_coverage,
     get_d4tools_coverage_completeness,
     get_d4tools_intervals_coverage,
     get_d4tools_intervals_mean_coverage,
     get_sample_interval_coverage,
     get_samples_sex_metrics,
-    set_interval,
 )
 from chanjo2.meta.handle_tasks import coverage_completeness_multitasker
 from chanjo2.models import SQLExon, SQLGene, SQLTranscript
@@ -55,8 +53,8 @@ def d4_interval_coverage(query: FileCoverageQuery):
     if None in [query.start, query.end]:  # Coverage over an entire chromosome
         return IntervalCoverage(
             mean_coverage=get_d4tools_chromosome_mean_coverage(
-                d4_file_path=query.coverage_file_path, chromosome=query.chromosome
-            ),
+                d4_file_path=query.coverage_file_path, chromosomes=[query.chromosome]
+            )[0][1],
             completeness={},
             interval_id=interval,
         )
@@ -135,14 +133,16 @@ def d4_intervals_coverage(query: FileCoverageIntervalsFileQuery):
 
 @router.get("/coverage/samples/predicted_sex", response_model=Dict)
 async def get_samples_predicted_sex(coverage_file_path: str):
-    try:
-        d4_file: D4File = get_d4_file(coverage_file_path=coverage_file_path)
-    except Exception:
+    """Return predicted sex for a sample given the coverage over its sex chromosomes."""
+    if (
+        isfile(coverage_file_path) is False
+        or validators.url(coverage_file_path) is False
+    ):
         raise HTTPException(
             status_code=status.HTTP_404_NOT_FOUND,
             detail=WRONG_COVERAGE_FILE_MSG,
         )
-    return get_samples_sex_metrics(d4_file=d4_file)
+    return get_samples_sex_metrics(d4_file_path=coverage_file_path)
 
 
 @router.post(

src/chanjo2/meta/handle_d4.py

@@ -46,18 +46,23 @@ def get_intervals_coords_list(
     return interval_coords
 
 
-def get_d4tools_chromosome_mean_coverage(d4_file_path: str, chromosome=str) -> float:
-    """Return mean coverage over one entire chromosome."""
+def get_d4tools_chromosome_mean_coverage(
+    d4_file_path: str, chromosomes=List[str]
+) -> List[Tuple[str, float]]:
+    """Return mean coverage over entire chromosomes."""
 
     chromosomes_stats_mean_cmd: List[str] = subprocess.check_output(
         ["d4tools", "stat", "-s" "mean", d4_file_path],
         text=True,
     ).splitlines()
-
+    chromosomes_coverage: List[Tuple[str, float]] = []
     for line in chromosomes_stats_mean_cmd:
         stats_data: List[str] = line.split("\t")
-        if chromosome == stats_data[CHROM_INDEX]:
-            return stats_data[STATS_MEAN_COVERAGE_INDEX]
+        if stats_data[CHROM_INDEX] in chromosomes:
+            chromosomes_coverage.append(
+                (stats_data[CHROM_INDEX], float(stats_data[STATS_MEAN_COVERAGE_INDEX]))
+            )
+    return chromosomes_coverage
 
 
 def get_d4tools_intervals_mean_coverage(
@@ -342,16 +347,17 @@ def predict_sex(x_cov: float, y_cov: float) -> str:
             return Sex.FEMALE.value
 
 
-def get_samples_sex_metrics(d4_file: D4File) -> Dict:
+def get_samples_sex_metrics(d4_file_path: str) -> Dict:
     """Compute coverage over sex chromosomes and predicted sex."""
 
-    sex_chroms_coverage: List[float] = get_intervals_mean_coverage(
-        d4_file=d4_file, intervals=[("X"), ("Y")]
+    sex_chroms_coverage: List[Tuple[str, float]] = get_d4tools_chromosome_mean_coverage(
+        d4_file_path=d4_file_path, chromosomes=["X", "Y"]
     )
+
     return {
-        "x_coverage": round(sex_chroms_coverage[0], 1),
-        "y_coverage": round(sex_chroms_coverage[1], 1),
+        "x_coverage": round(sex_chroms_coverage[0][1], 1),
+        "y_coverage": round(sex_chroms_coverage[1][1], 1),
         "predicted_sex": predict_sex(
-            x_cov=sex_chroms_coverage[0], y_cov=sex_chroms_coverage[1]
+            x_cov=sex_chroms_coverage[0][1], y_cov=sex_chroms_coverage[1][1]
         ),
     }

src/chanjo2/meta/handle_report_contents.py

@@ -113,9 +113,7 @@ def get_report_data(
         return data
 
     # Add coverage_report - specific data
-    data["sex_rows"] = get_report_sex_rows(
-        samples=query.samples, samples_d4_files=samples_d4_files
-    )
+    data["sex_rows"] = get_report_sex_rows(samples=query.samples)
     data["completeness_rows"] = get_report_completeness_rows(
         samples_coverage_stats=samples_coverage_stats,
         levels=query.completeness_thresholds,
@@ -231,17 +229,13 @@ def get_report_completeness_rows(
     return completeness_rows
 
 
-def get_report_sex_rows(
-    samples: List[ReportQuerySample], samples_d4_files: List[Tuple[str, D4File]]
-) -> List[Dict]:
+def get_report_sex_rows(samples: List[ReportQuerySample]) -> List[Dict]:
     """Create and return the contents for the sample sex lines in the coverage report."""
     sample_sex_rows: D4FileList = []
     for sample in samples:
-        for identifier, d4_file in samples_d4_files:
-            if identifier != sample.name:
-                continue
-
-        sample_sex_metrics: Dict = get_samples_sex_metrics(d4_file=d4_file)
+        sample_sex_metrics: Dict = get_samples_sex_metrics(
+            d4_file_path=sample.coverage_file_path
+        )
 
         sample_sex_row: SampleSexRow = SampleSexRow(
             **{