Tested and completed adding call sample filtering

.vscode/settings.json

@@ -10,8 +10,10 @@
         "src"
     ],
     "cSpell.words": [
+        "CHROM",
         "pytest",
         "pyvcf",
+        "vcfpy",
         "vcfs"
     ],
 }

src/modvcfsamples/cli.py

@@ -4,9 +4,9 @@ from modvcfsamples import sample
 
 def run(vcfs: list[str], only: list[str], output_dir: str):
     for vcf in vcfs:
-        vcf_records = sample.get_records_from_vcf(vcf)
-        sample.filter_all_sample_datatypes(vcf_records, *only)
-        sample.write_records_to_vcf(vcf_records, os.path.join(output_dir, os.path.basename(vcf)))
+        vcf_records, header = sample.get_records_from_vcf(vcf)
+        processed_vcfs, modified_header = sample.keep_specific_call_data(vcf_records, header, *only)
+        sample.write_records_to_vcf(processed_vcfs, modified_header, os.path.join(output_dir, os.path.basename(vcf)))
 
 def main():
     parser = argparse.ArgumentParser()

src/modvcfsamples/sample.py

@@ -1,34 +1,58 @@
-import vcf
+from typing import Iterable, OrderedDict
+import vcfpy
 from collections import namedtuple
 from copy import deepcopy
+import os
+
+from vcfpy.record import Call
+
 
 def get_records_from_vcf(path: str):
     vcf_records = []
     with open(path, "r") as vcf_stream:
-        reader = vcf.Reader(vcf_stream)
+        reader = vcfpy.Reader.from_stream(vcf_stream)
         for record in reader:
             vcf_records.append(record)
 
-    return vcf_records
+    return vcf_records, reader.header
 
-def filter_sample_datatype(record, *datatypes: str):
-    call_data = namedtuple("Data", *datatypes)
-    filtered_calls = []
-    modified_record = deepcopy(record)
-    for call in record.samples:
-        kept_data = {}
-        for datatype in datatypes:
-            kept_data[datatype] = call[datatype]
-        filtered_calls.append(call_data(**kept_data))
-    modified_record.samples = filtered_calls
-    return modified_record
 
-def filter_all_sample_datatypes(records: list, *datatypes: str):
+def keep_specific_call_data(records: list[vcfpy.Record], header: vcfpy.Header, *datatypes: str):
+    lines_kept = [line for line in header.lines if line.key != "FORMAT"]
+    lines_kept.extend([line for line in header.get_lines("FORMAT") if line.id in datatypes])
+    modified_header = vcfpy.Header(lines_kept, header.samples)
+    modified_records = []
+
     for record in records:
-        yield filter_sample_datatype(record, *datatypes)
+        modified_calls = []
+        for call in record.calls:
+            kept_data = OrderedDict()
+            for datatype in datatypes:
+                kept_data[datatype] = call.data[datatype]
+            modified_calls.append(Call(call.sample, kept_data, record))
+        modified_record = vcfpy.Record(
+            record.CHROM,
+            record.POS,
+            record.ID,
+            record.REF,
+            record.ALT,
+            record.QUAL,
+            record.FILTER,
+            record.INFO,
+            datatypes,
+            modified_calls,
+        )
+        modified_records.append(modified_record)
+    return modified_records, modified_header
 
-def write_records_to_vcf(records: list, path: str):
+def normalize_gt_to_length(records: list[vcfpy.Record], header: vcfpy.Header, num: int):
+    for record in records:
+        for call in record.calls:
+            pass
+
+def write_records_to_vcf(records: Iterable[vcfpy.Record], header: vcfpy.Header, path: str):
+    os.makedirs(os.path.dirname(path), exist_ok=True)
     with open(path, "w") as vcf_stream:
-        writer = vcf.Writer(vcf_stream, records[0])
+        writer = vcfpy.Writer.from_stream(vcf_stream, header)
         for record in records:
-            writer.write_record(record)
+            writer.write_record(record)

tests/modvcfsamples/test_sample.py

@@ -1,38 +1,18 @@
-from modvcfsamples.sample import filter_sample_datatype, filter_all_sample_datatypes, get_records_from_vcf
+from modvcfsamples.sample import keep_specific_call_data, get_records_from_vcf
 import os
 
-def test_get_records_from_vcf_not_none():
-    records = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
-    assert len(records) > 0
-
-def test_filter_sample_datatype_not_none():
-    records = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
-    filter_for = ["GT"]
-    modified_record = filter_sample_datatype(records[0], *filter_for)
-    assert modified_record is not None
-
-def test_filter_sample_datatype_only_filtered():
-    records = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
-    filter_for = ["GT"]
-    modified_record = filter_sample_datatype(records[0], *filter_for)
-    for sample in modified_record.samples:
-        assert len(sample) <= len(filter_for)
-        for key, _ in sample._asdict().items():
-            assert key in filter_for
-
 def test_filter_all_sample_datatypes_not_empty():
-    records = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
+    records, header = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
     filter_for = ["GT"]
-    modified_records = list(filter_all_sample_datatypes(records, *filter_for))
+    modified_records, header = keep_specific_call_data(records, header, *filter_for)
     assert len(modified_records) == 11
 
 def test_filter_all_sample_datatypes_filtered():
-    records = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
+    records, header = get_records_from_vcf(os.path.abspath("tests/resources/test_files_shortened.vcf"))
     filter_for = ["GT"]
-    modified_records = list(filter_all_sample_datatypes(records, *filter_for))
+    modified_records, header = keep_specific_call_data(records, header, *filter_for)
     for modified_record in modified_records:
-        for sample in modified_record.samples:
-            assert len(sample) <= len(filter_for)
-            for key, _ in sample._asdict().items():
+        for call in modified_record.calls:
+            assert len(call.data.keys()) <= len(filter_for)
+            for key, _ in call.data.items():
                 assert key in filter_for
-

tests/resources/expected_out/test_files_shortened.vcf

@@ -0,0 +1,17 @@
+##fileformat=VCFv4.1
+##fileDate=10122015_22h01m13s
+##source=SHAPEIT2.v837
+##log_file=shapeit_10122015_22h01m13s_3f764d75-2fbb-42df-ab75-8c2dfd5731ce.log
+##FORMAT=<ID=GT,Number=1,Type=String,Description="Phased Genotype">
+#CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	Gambian	Chinese	French	Brazilian	Nigerian	Pakistani	English	Colombian	Indian	Japanese
+1	846808	rs4475691	C	T	100	.	AC=1276;AN=5008;DP=20368	GT	1|1	0|0	0|0	./.	0|0	0|0	0|0	0|0	./.	0|0
+1	846854	rs111957712	G	A	100	.	AC=114;AN=5008;DP=20538	GT	./.	0|0	0|0	0|0	0|0	0|0	./.	./.	0|0	0|0
+1	846864	rs950122	G	C	100	.	AC=1116;AN=5008;DP=20582	GT	1|1	0|0	0|0	0|0	0|0	0|0	0|0	0|0	./.	0|0
+1	847228	rs3905286	C	T	100	.	AC=1215;AN=5008;DP=20731	GT	1|1	0|0	0|0	0|0	0|0	0|0	0|0	0|0	0|1	0|0
+1	847297	rs11507768	G	A	100	.	AC=359;AN=5008;DP=20809	GT	1|0	0|0	0|0	./.	0|0	0|0	./.	0|0	0|0	0|0
+1	847491	rs28407778	G	A	100	.	AC=1262;AN=5008;DP=16939	GT	1|1	0|0	0|0	0|0	0|0	./.	0|0	0|0	0|1	0|0
+1	848023	rs144407116	C	A	100	.	AC=52;AN=5008;DP=22562	GT	0|1	0|0	0|0	0|0	0|0	0|0	0|0	0|0	0|0	0|0
+1	848090	rs4246505	G	A	100	.	AC=857;AN=5008;DP=19301	GT	0|0	0|0	0|0	0|0	0|0	0|0	./.	0|0	./.	0|0
+1	848445	rs4626817	G	A	100	.	AC=1255;AN=5008;DP=18444	GT	1|1	0|0	0|0	0|0	0|0	0|0	0|0	./.	./.	0|0
+1	848456	rs11507767	A	G	100	.	AC=1266;AN=5008;DP=18137	GT	1|1	./.	0|0	0|0	0|0	0|0	0|0	0|0	0|1	./.
+1	848738	rs3829741	C	T	100	.	AC=855;AN=5008;DP=16663	GT	0|0	0|0	0|0	0|0	0|0	0|0	0|0	0|0	0|1	0|0