Repository 'cravat_vcf_convert'
hg clone https://toolshed.g2.bx.psu.edu/repos/in_silico/cravat_vcf_convert

Changeset 2:c042835a7163 (2018-06-12)
Previous changeset 1:152cc8feff8a (2018-06-12) Next changeset 3:b968ba302ba6 (2018-06-12)
Commit message:
Uploaded
added:
cravat_convert/vcf_converter.py
removed:
cravat_convert/cravat_convert.py
cravat_convert/cravat_convert.xml
b
diff -r 152cc8feff8a -r c042835a7163 cravat_convert/cravat_convert.py
--- a/cravat_convert/cravat_convert.py Tue Jun 12 11:05:27 2018 -0400
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
[
@@ -1,80 +0,0 @@
-'''
-Convert a VCF format file to Cravat format file
-'''
-
-import os
-import argparse
-from vcf_converter import CravatConverter
-###
-import ipdb
-
-
-# File read/write configuration variables
-vcf_sep = '\t'
-cr_sep = '\t'
-cr_newline = '\n'
-
-# VCF Headers mapped to their index position in a row of VCF values
-vcf_mapping = {
-    'CHROM': 0,
-    'POS': 1,
-    'ID': 2,
-    'REF': 3,
-    'ALT': 4,
-    'QUAL': 5,
-    'FILTER': 6,
-    'INFO': 7,
-    'FORMAT': 8,
-    'NA00001': 9,
-    'NA00002': 10,
-    'NA00003': 11
-}
-
-
-def get_args():
-    parser = argparse.ArgumentParser()
-    parser.add_argument('--input',
-                            '-i',
-                            required = True,
-                            help='Input path to a VCF file for conversion',)
-    parser.add_argument('--output',
-                            '-o',
-                            default = os.path.join(os.getcwd(), "cravat_converted.txt"),
-                            help = 'Output path to write the cravat file to')
-    return parser.parse_args()
-
-
-def convert(in_path, out_path=None):
-    if not out_path:
-        base, _ = os.path.split(in_path)
-        out_path = os.path.join(base, "cravat_converted.txt")
-    
-    with open(in_path, 'r') as in_file, \
-    open(out_path, 'w') as out_file:
-
-        # cr_count will be used to generate the 'TR' field of the cravat rows (first header)
-        cr_count = 0
-        # VCF lines are always assumed to be '+' strand, as VCF doesn't specify that attribute
-        strand = '+'
-        # VCF converter. Adjusts position, reference, and alternate for Cravat formatting.
-        converter = CravatConverter()
-
-        for line in in_file:
-            if line.startswith("#"):
-                continue
-            line = line.strip().split(vcf_sep)
-            # row is dict of VCF headers mapped to corresponding values of this line
-            row = { header: line[index] for header, index in vcf_mapping.items() }
-            for alt in row["ALT"].split(","):
-                new_pos, new_ref, new_alt = converter.extract_vcf_variant(strand, row["POS"], row["REF"], alt)
-                new_pos, new_ref, new_alt = str(new_pos), str(new_ref), str(new_alt)
-                cr_line = cr_sep.join([
-                    'TR' + str(cr_count), row['CHROM'], new_pos, strand, new_ref, new_alt, row['ID']
-                ])
-                out_file.write(cr_line + cr_newline)
-                cr_count += 1
-
-
-if __name__ == "__main__":
-    cli_args = get_args()
-    convert(cli_args.input, cli_args.output)
b
diff -r 152cc8feff8a -r c042835a7163 cravat_convert/cravat_convert.xml
--- a/cravat_convert/cravat_convert.xml Tue Jun 12 11:05:27 2018 -0400
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
b
@@ -1,20 +0,0 @@
-<tool id="cravat_convert" name="CRAVAT Convert" version="1.0.0">
-    <description>Converts a VCF format file to a Cravat format file</description>
-    <command interpreter="python">cravat_convert.py -i $input -o $output</command>
-  
-    <inputs>
-        <param format="tabular" name="input" type="data" label="Source file"/>
-    </inputs>
-  
-    <outputs>
-        <data format="tabular" name="output" />
-    </outputs>
-
-    <!-- <tests></tests> -->
-
-    <help>
-        Converts a VCF format file to a Cravat format file
-    </help>
-
-</tool>
-
b
diff -r 152cc8feff8a -r c042835a7163 cravat_convert/vcf_converter.py
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/cravat_convert/vcf_converter.py Tue Jun 12 11:20:00 2018 -0400
[
b'@@ -0,0 +1,243 @@\n+"""\r\n+A module originally obtained from the cravat package. Modified to use in the vcf\r\n+converter galaxy tool.\r\n+\r\n+\r\n+Register of changes made (Chris Jacoby):\r\n+    1) Changed imports as galaxy tool won\'t have access to complete cravat python package\r\n+    2) Defined BadFormatError in BaseConverted file, as I didn\'t have the BadFormatError module\r\n+"""\r\n+\r\n+from base_converter import BaseConverter, BadFormatError\r\n+import re\r\n+\r\n+class CravatConverter(BaseConverter):\r\n+    \r\n+    def __init__(self):\r\n+        self.format_name = \'vcf\'\r\n+        self.samples = []\r\n+        self.var_counter = 0\r\n+        self.addl_cols = [{\'name\':\'phred\',\r\n+                           \'title\':\'Phred\',\r\n+                           \'type\':\'string\'},\r\n+                          {\'name\':\'filter\',\r\n+                           \'title\':\'VCF filter\',\r\n+                           \'type\':\'string\'},\r\n+                          {\'name\':\'zygosity\',\r\n+                           \'title\':\'Zygosity\',\r\n+                           \'type\':\'string\'},\r\n+                          {\'name\':\'alt_reads\',\r\n+                           \'title\':\'Alternate reads\',\r\n+                           \'type\':\'int\'},\r\n+                          {\'name\':\'tot_reads\',\r\n+                           \'title\':\'Total reads\',\r\n+                           \'type\':\'int\'},\r\n+                          {\'name\':\'af\',\r\n+                           \'title\':\'Variant allele frequency\',\r\n+                           \'type\':\'float\'}]\r\n+    \r\n+    def check_format(self, f): \r\n+        return f.readline().startswith(\'##fileformat=VCF\')\r\n+    \r\n+    def setup(self, f):\r\n+        \r\n+        vcf_line_no = 0\r\n+        for line in f:\r\n+            vcf_line_no += 1\r\n+            if len(line) < 6:\r\n+                continue\r\n+            if line[:6] == \'#CHROM\':\r\n+                toks = re.split(\'\\s+\', line.rstrip())\r\n+                if len(toks) > 8:\r\n+                    self.samples = toks[9:]\r\n+                break\r\n+    \r\n+    def convert_line(self, l):\r\n+        if l.startswith(\'#\'): return None\r\n+        self.var_counter += 1\r\n+        toks = l.strip(\'\\r\\n\').split(\'\\t\')\r\n+        all_wdicts = []\r\n+        if len(toks) < 8:\r\n+            raise BadFormatError(\'Wrong VCF format\')\r\n+        [chrom, pos, tag, ref, alts, qual, filter, info] = toks[:8]\r\n+        if tag == \'\':\r\n+            raise BadFormatError(\'ID column is blank\')\r\n+        elif tag == \'.\':\r\n+            tag = \'VAR\' + str(self.var_counter)\r\n+        if chrom[:3] != \'chr\':\r\n+            chrom = \'chr\' + chrom\r\n+        alts = alts.split(\',\')\r\n+        len_alts = len(alts)\r\n+        if len(toks) == 8:\r\n+            for altno in range(len_alts):\r\n+                wdict = None\r\n+                alt = alts[altno]\r\n+                newpos, newref, newalt = self.extract_vcf_variant(\'+\', pos, ref, alt)\r\n+                wdict = {\'tags\':tag,\r\n+                         \'chrom\':chrom,\r\n+                         \'pos\':newpos,\r\n+                         \'ref_base\':newref,\r\n+                         \'alt_base\':newalt,\r\n+                         \'sample_id\':\'no_sample\',\r\n+                         \'phred\': qual,\r\n+                         \'filter\': filter}\r\n+                all_wdicts.append(wdict)\r\n+        elif len(toks) > 8:\r\n+            sample_datas = toks[9:]\r\n+            genotype_fields = {}\r\n+            genotype_field_no = 0\r\n+            for genotype_field in toks[8].split(\':\'):\r\n+                genotype_fields[genotype_field] = genotype_field_no\r\n+                genotype_field_no += 1\r\n+            if not (\'GT\' in genotype_fields):\r\n+                raise BadFormatError(\'No GT Field\')\r\n+            gt_field_no = genotype_fields[\'GT\']\r\n+            for sample_no in range(len(sample_datas)):\r\n+                sample = self.samples[sample_no]\r\n+                sample_data = sample_datas[sample_no].split(\':\')\r\n+                gts = {}\r\n+                for gt in sample_data[gt_field_no].replace(\'/\', \'|\').split(\'|\'):\r\n+                  '..b'\r\n+                ref_reads = sample_data[genotype_fields[\'AD\']].split(\',\')[0]\r\n+                alt_reads = sample_data[genotype_fields[\'AD\']].split(\',\')[1]\r\n+            elif gt == max(gts.keys()):    \r\n+                #if geontype has multiple alt bases, then AD will have #alt1 reads, #alt2 reads\r\n+                alt_reads = sample_data[genotype_fields[\'AD\']].split(\',\')[1]\r\n+            else:\r\n+                alt_reads = sample_data[genotype_fields[\'AD\']].split(\',\')[0]                            \r\n+                             \r\n+        if \'DP\' in genotype_fields and genotype_fields[\'DP\'] <= len(sample_data): \r\n+            depth = sample_data[genotype_fields[\'DP\']] \r\n+        elif alt_reads != \'\' and ref_reads != \'\':\r\n+            #if DP is not present but we have alt and ref reads count, dp = ref+alt\r\n+            depth = int(alt_reads) + int(ref_reads)   \r\n+\r\n+        if \'AF\' in genotype_fields and genotype_fields[\'AF\'] <= len(sample_data):\r\n+            af = float(sample_data[genotype_fields[\'AF\']] )\r\n+        elif depth != \'\' and alt_reads != \'\':\r\n+            #if AF not specified, calc it from alt and ref reads\r\n+            af = float(alt_reads) / float(depth)\r\n+ \r\n+        return depth, alt_reads, af\r\n+            \r\n+    def extract_vcf_variant (self, strand, pos, ref, alt):\r\n+\r\n+        reflen = len(ref)\r\n+        altlen = len(alt)\r\n+        \r\n+        # Returns without change if same single nucleotide for ref and alt. \r\n+        if reflen == 1 and altlen == 1 and ref == alt:\r\n+            return pos, ref, alt\r\n+        \r\n+        # Trimming from the start and then the end of the sequence \r\n+        # where the sequences overlap with the same nucleotides\r\n+        new_ref2, new_alt2, new_pos = \\\r\n+            self.trimming_vcf_input(ref, alt, pos, strand)\r\n+                \r\n+        if new_ref2 == \'\':\r\n+            new_ref2 = \'-\'\r\n+        if new_alt2 == \'\':\r\n+            new_alt2 = \'-\'\r\n+        \r\n+        return new_pos, new_ref2, new_alt2\r\n+    \r\n+    # This function looks at the ref and alt sequences and removes \r\n+    # where the overlapping sequences contain the same nucleotide.\r\n+    # This trims from the end first but does not remove the first nucleotide \r\n+    # because based on the format of VCF input the \r\n+    # first nucleotide of the ref and alt sequence occur \r\n+    # at the position specified.\r\n+    #     End removed first, not the first nucleotide\r\n+    #     Front removed and position changed\r\n+    def trimming_vcf_input(self, ref, alt, pos, strand):\r\n+        pos = int(pos)\r\n+        reflen = len(ref)\r\n+        altlen = len(alt)\r\n+        minlen = min(reflen, altlen)\r\n+        new_ref = ref\r\n+        new_alt = alt\r\n+        new_pos = pos\r\n+        # Trims from the end. Except don\'t remove the first nucleotide. \r\n+        # 1:6530968 CTCA -> GTCTCA becomes C -> GTC.\r\n+        for nt_pos in range(0, minlen - 1): \r\n+            if ref[reflen - nt_pos - 1] == alt[altlen - nt_pos - 1]:\r\n+                new_ref = ref[:reflen - nt_pos - 1]\r\n+                new_alt = alt[:altlen - nt_pos - 1]\r\n+            else:\r\n+                break    \r\n+        new_ref_len = len(new_ref)\r\n+        new_alt_len = len(new_alt)\r\n+        minlen = min(new_ref_len, new_alt_len)\r\n+        new_ref2 = new_ref\r\n+        new_alt2 = new_alt\r\n+        # Trims from the start. 1:6530968 G -> GT becomes 1:6530969 - -> T.\r\n+        for nt_pos in range(0, minlen):\r\n+            if new_ref[nt_pos] == new_alt[nt_pos]:\r\n+                if strand == \'+\':\r\n+                    new_pos += 1\r\n+                elif strand == \'-\':\r\n+                    new_pos -= 1\r\n+                new_ref2 = new_ref[nt_pos + 1:]\r\n+                new_alt2 = new_alt[nt_pos + 1:]\r\n+            else:\r\n+                new_ref2 = new_ref[nt_pos:]\r\n+                new_alt2 = new_alt[nt_pos:]\r\n+                break  \r\n+        return new_ref2, new_alt2, new_pos\r\n+\r\n+\r\n+if __name__ == "__main__":\r\n+    c = CravatConverter()\n\\ No newline at end of file\n'