Merge pull request #3 from NCI-GDC/feature/binf-687-modify-sample

[BINF-687] Add modify vcf sample script

.pre-commit-config.yaml

@@ -3,13 +3,13 @@ default_language_version:
   python_venv: python3.8
 
 repos:
-  - repo: git@github.com:Yelp/detect-secrets
-    rev: v1.0.3
+  - repo: https://github.com/Yelp/detect-secrets
+    rev: v1.2.0
     hooks:
       - id: detect-secrets
         args: ['--baseline', '.secrets.baseline']
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v2.4.0
+    rev: v4.3.0
     hooks:
       - id: check-yaml
       - id: check-toml
@@ -18,16 +18,16 @@ repos:
         args: ["--allow-missing-credentials"]
       - id: detect-private-key
   - repo: https://github.com/pycqa/isort
-    rev: 5.8.0
+    rev: 5.10.1
     hooks:
       - id: isort
         name: isort
   - repo: https://github.com/psf/black
-    rev: 20.8b1
+    rev: 22.3.0
     hooks:
       - id: black
   - repo: https://github.com/pycqa/flake8
-    rev: '3.9.2'
+    rev: '4.0.1'
     hooks:
     -   id: flake8
         additional_dependencies: [flake8-docstrings]

.secrets.baseline

@@ -1,5 +1,5 @@
 {
-  "version": "1.0.3",
+  "version": "1.2.0",
   "plugins_used": [
     {
       "name": "ArtifactoryDetector"
@@ -97,6 +97,15 @@
       "pattern": [
         "^.secrets.baseline$"
       ]
+    },
+    {
+      "path": "detect_secrets.filters.heuristic.is_lock_file"
+    },
+    {
+      "path": "detect_secrets.filters.heuristic.is_not_alphanumeric_string"
+    },
+    {
+      "path": "detect_secrets.filters.heuristic.is_swagger_file"
     }
   ],
   "results": {

Dockerfile

@@ -15,6 +15,12 @@ COPY requirements.txt /opt/dist
 
 WORKDIR /opt/dist
 
+RUN apt update -y \
+	&& apt install -y \
+		libbz2-dev \
+		liblzma-dev \
+		zlib1g
+
 RUN pip install -r requirements.txt \
 	&& pip install *.tar.gz \
 	&& rm -f *.tar.gz requirements.txt

dev-requirements.txt

@@ -18,8 +18,10 @@ cfgv==3.1.0
     # via pre-commit
 chardet==3.0.4
     # via requests
-click==8.0.1
-    # via black
+click==8.1.3
+    # via
+    #   -c requirements.txt
+    #   black
 coverage==5.2.1
     # via pytest-cov
 detect-secrets==1.0.3

gatk4_mutect2_tool/modify_vcf_sample.py

@@ -0,0 +1,93 @@
+#!/usr/bin/env python3
+"""
+Modify GATK4.1.2 Mutect2 VCF sample header to "TUMOR", "NORMAL"
+
+@author: Shenglai Li
+"""
+import logging
+from typing import Optional
+
+import click
+import pysam
+
+logger = logging.getLogger(__name__)
+logging.basicConfig(level=logging.INFO)
+
+tumor_bamT = str
+normal_bamT = Optional[str]
+vcfT = str
+outputT = str
+
+
+def get_sample_name(bam: str) -> str:
+    '''
+    Get sample name from BAM file
+    '''
+    sample: str
+    b = pysam.AlignmentFile(bam, 'rb')
+    sample = b.header['RG'][0]['SM']  # type: ignore
+    return sample
+
+
+def modify_vcf_sample(
+    tumor_bam: tumor_bamT, normal_bam: normal_bamT, vcf: vcfT, output: outputT
+) -> None:
+    '''
+    Modify VCF sample in the header
+    '''
+    out_vcf = output
+    reader = pysam.BGZFile(vcf, mode='rb')  # type: ignore
+    writer = pysam.BGZFile(out_vcf, mode='wb')  # type: ignore
+    try:
+        for line in reader:
+            new_line: str
+            new_line = line.decode('utf-8')
+            if new_line.startswith('#CHROM'):
+                if normal_bam:
+                    assert (
+                        normal_bam in new_line
+                    ), f'Unable to find normal sample tag in the vcf file. {normal_bam}'
+                    new_line = new_line.replace(f'{normal_bam}', 'NORMAL')
+                assert (
+                    tumor_bam in new_line
+                ), f'Unable to find tumor sample tag in the vcf file. {tumor_bam}'
+                new_line = new_line.replace(f'{tumor_bam}', 'TUMOR')
+                writer.write(str.encode(f"{new_line}\n", encoding='utf-8'))
+            else:
+                new_line = new_line + '\n'
+                writer.write(new_line.encode('utf-8'))
+    except AssertionError as e:
+        logger.exception(e)
+    finally:
+        writer.close()
+        reader.close()
+    pysam.tabix_index(out_vcf, preset='vcf', force=True)
+
+
+@click.command()
+@click.option('--tumor_bam', required=True)
+@click.option('--vcf', required=True)
+@click.option('--output', required=True)
+@click.option('--normal_bam', required=False)
+def main(
+    tumor_bam: tumor_bamT,
+    vcf: vcfT,
+    output: outputT,
+    normal_bam: normal_bamT = None,
+) -> None:
+    '''
+    main
+    '''
+    tumor_sample = get_sample_name(tumor_bam)
+    logger.info(f'{tumor_sample=}')
+    normal_sample = None
+    if normal_bam:
+        normal_sample = get_sample_name(normal_bam)
+    logger.info(f'{normal_sample=}')
+    modify_vcf_sample(tumor_sample, normal_sample, vcf, output)
+
+
+if __name__ == "__main__":
+    main()
+
+# __END__

pyproject.toml

@@ -25,3 +25,9 @@ python_version = 3.8
 disallow_untyped_defs = true
 warn_return_any = true
 warn_unused_configs = true
+
+[[tool.mypy.overrides]]
+module = [
+        "pysam"
+]
+ignore_missing_imports = true

requirements.txt

@@ -4,3 +4,7 @@
 #
 #    pip-compile --output-file=requirements.txt
 #
+click==8.1.3
+    # via gatk4-mutect2-tool (setup.py)
+pysam==0.19.1
+    # via gatk4-mutect2-tool (setup.py)

setup.cfg

@@ -15,6 +15,8 @@ classifiers =
 
 [options]
 install_requires =
+        click
+        pysam
 zip_safe = False
 include_package_data = True
 packages = find:
@@ -25,6 +27,7 @@ setup_requires =
 [options.entry_points]
 console_scripts =
         gatk4_mutect2_tool = gatk4_mutect2_tool.__main__:main
+        modify_vcf_sample = gatk4_mutect2_tool.modify_vcf_sample:main
 
 [coverage:run]
 branch = true

tests/test_modify_vcf_sample.py

@@ -0,0 +1,32 @@
+#!/usr/bin/env python3
+
+import unittest
+
+from click.testing import CliRunner
+
+from gatk4_mutect2_tool import modify_vcf_sample as MOD
+
+
+class ThisTestCase(unittest.TestCase):
+    def setUp(self) -> None:
+        super().setUp()
+        self.cli_args_dict = {
+            "--normal_bam": "normal.bam",
+            "--tumor_bam": "tumor.bam",
+            "--vcf": "test.vcf",
+            "--output": "output.vcf",
+        }
+        self.cli_args_list = []
+        for k, v in self.cli_args_dict.items():
+            self.cli_args_list.extend([k, v])
+
+    def cli_args(self):
+        runner = CliRunner()
+        result = runner.invoke(MOD.main, self.cli_args_list)
+        self.assertEqual(result.exit_code, 0)
+
+
+if __name__ == "__main__":
+    unittest.main()
+
+# __END__