python/vcf/str1.py

   1 #!/usr/bin/env python3
   2
   3 import sys
   4 import os
   5 from pathlib import Path
   6 from pysam import VariantFile
   7
   8 import pprint
   9
  10 def readVcfDat(fname):
  11     vcf_in = VariantFile(fname)
  12     for rec in vcf_in.fetch():
  13         pprint.pprint(rec)
  14         print ([rec.chrom,rec.pos,rec.id,rec.alleles])
  15         if rec.id.startswith('rs'):
  16             # SNP
  17         else:
  18             # STR
  19         for key,value in rec.samples.iteritems():
  20             print((key, value['GT']))
  21
  22
  23 def main():
  24     try:
  25         p = Path('unphased_all_vcf')
  26         for f in p.iterdir():
  27             if f.is_file():
  28                 if '.vcf' in f.suffixes:
  29                     print((f,f.suffixes))
  30                     readVcfDat(f)
  31
  32         print ('[!]Done', file=sys.stderr)
  33         sys.stdout.flush()
  34     except BrokenPipeError: # https://stackoverflow.com/a/58517082
  35         # https://docs.python.org/3/library/signal.html#note-on-sigpipe
  36         # Python flushes standard streams on exit; redirect remaining output
  37         # to devnull to avoid another BrokenPipeError at shutdown
  38         devnull = os.open(os.devnull, os.O_WRONLY)
  39         os.dup2(devnull, sys.stdout.fileno())
  40         sys.exit(1)  # Python exits with error code 1 on EPIPE
  41
  42 if __name__ == '__main__':
  43     main()