Merge pull request #2 from amplab/master

off by one error in writing normalize variants
amplab · May 14, 2014 · b414114 · b414114
2 parents e926574 + 3757bd3
commit b414114
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 5 deletions.
diff --git a/smashbenchmarking/normalize_vcf.py b/smashbenchmarking/normalize_vcf.py
@@ -78,10 +78,9 @@ def genotype(vcfrecord):
         return vcfrecord.samples[0].gt_nums
     return {0 : "0/0", 1 : "0/1", 2: "1/1", None : "."}[vcfrecord.samples[0].gt_type]
 
-# assume that input position is 0-based, so offset by 1
 def write(record, writer):
-    return writer.write_record(record.CHROM, record.POS + 1, '.',
-                               record.REF, record.ALT, genotype(record)) # TODO: more gtypes.
+    return writer.write_record(record.CHROM, record.POS, '.',
+                               record.REF, ','.join(map(lambda a: str(a),record.ALT)), genotype(record)) # TODO: more gtypes.
 
 
 left_slides = []

diff --git a/smashbenchmarking/parsers/vcfwriter.py b/smashbenchmarking/parsers/vcfwriter.py
@@ -84,7 +84,7 @@ def write_record(self, CHROM, POS, ID, REF, ALT, gtype):
             QUAL = 20               # Default 1/100 error probability.
             FILTER = 'PASS'
             FORMAT = 'GT'
-            print(CHROM, POS + 1, ID, REF, ALT, QUAL, FILTER, '.', FORMAT, gtype,
+            print(CHROM, POS, ID, REF, ALT, QUAL, FILTER, '.', FORMAT, gtype,
                   sep='\t', file=self._output)
         return write
 

diff --git a/test/normalize.py b/test/normalize.py
@@ -51,7 +51,7 @@ def getVcf(self,str):
         vcf_io = StringIO.StringIO(str)
         return vcf.Reader(vcf_io)
 
-    def normalizeString(self,vcf_str):
+    def normalizeStringToWriter(self,vcf_str):
         vcf_io = StringIO.StringIO(vcf_str)
         test_vcf = vcf.Reader(vcf_io)
         output_io = StringIO.StringIO()
@@ -234,5 +234,17 @@ def testMultipleAltAlleles(self):
         self.assertEqual(record.REF,'G')
         self.assertEqual(record.ALT[0],'CG')
 
+    def testNormalizerWriter(self):
+        vcf_str = """##fileformat=VCFv4.0
+##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">\n
+#CHROM  POS     ID      REF     ALT     QUAL    FILTER  INFO    FORMAT  NA00001\n
+chr1   2   .       a     c       20      PASS    .       GT      0/1\n
+chr1   4   .      A      G       20      PASS     .      GT      1/1\n
+"""
+        output_vcf = self.normalizeStringToWriter(vcf_str)
+        r1 = output_vcf.next()
+        self.assertEqual(r1.POS,2)
+
+
 if __name__ == '__main__':
     unittest.main()