Mercurial > repos > xuebing > sharplabtool

comparison tools/genome_diversity/extract_flanking_dna.py @ 0:9071e359b9a3

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author xuebing
date Fri, 09 Mar 2012 19:37:19 -0500
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:9071e359b9a3
1 #!/usr/bin/env python2.5
2
3 import os
4 import sys
5 from optparse import OptionParser
6 import genome_diversity as gd
7
8 def main_function( parse_arguments=None ):
9 if parse_arguments is None:
10 parse_arguments = lambda arguments: ( None, arguments )
11 def main_decorator( to_decorate ):
12 def decorated_main( arguments=None ):
13 if arguments is None:
14 arguments = sys.argv
15 options, arguments = parse_arguments( arguments )
16 rc = 1
17 try:
18 rc = to_decorate( options, arguments )
19 except Exception, err:
20 sys.stderr.write( 'ERROR: %s\n' % str( err ) )
21 traceback.print_exc()
22 finally:
23 sys.exit( rc )
24 return decorated_main
25 return main_decorator
26
27 def parse_arguments( arguments ):
28 parser = OptionParser()
29 parser.add_option('--input',
30 type='string', dest='input',
31 help='file of selected SNPs')
32 parser.add_option('--output',
33 type='string', dest='output',
34 help='output file')
35 parser.add_option('--snps_loc',
36 type='string', dest='snps_loc',
37 help='snps .loc file')
38 parser.add_option('--scaffold_col',
39 type="int", dest='scaffold_col',
40 help='scaffold column in the input file')
41 parser.add_option('--pos_col',
42 type="int", dest='pos_col',
43 help='position column in the input file')
44 parser.add_option('--output_format',
45 type="string", dest='output_format',
46 help='output format, fasta or primer3')
47 parser.add_option('--species',
48 type="string", dest='species',
49 help='species')
50 return parser.parse_args( arguments[1:] )
51
52
53 @main_function( parse_arguments )
54 def main( options, arguments ):
55 if not options.input:
56 raise RuntimeError( 'missing --input option' )
57 if not options.output:
58 raise RuntimeError( 'missing --output option' )
59 if not options.snps_loc:
60 raise RuntimeError( 'missing --snps_loc option' )
61 if not options.scaffold_col:
62 raise RuntimeError( 'missing --scaffold_col option' )
63 if not options.pos_col:
64 raise RuntimeError( 'missing --pos_col option' )
65 if not options.output_format:
66 raise RuntimeError( 'missing --output_format option' )
67 if not options.species:
68 raise RuntimeError( 'missing --species option' )
69
70 snps = gd.SnpFile( filename=options.input, seq_col=int( options.scaffold_col ), pos_col=int( options.pos_col ) )
71
72 out_fh = gd._openfile( options.output, 'w' )
73
74 snpcalls_file = gd.get_filename_from_loc( options.species, options.snps_loc )
75 file_root, file_ext = os.path.splitext( snpcalls_file )
76 snpcalls_index_file = file_root + ".cdb"
77 snpcalls = gd.SnpcallsFile( data_file=snpcalls_file, index_file=snpcalls_index_file )
78
79 while snps.next():
80 seq, pos = snps.get_seq_pos()
81 flanking_dna = snpcalls.get_flanking_dna( sequence=seq, position=pos, format=options.output_format )
82 if flanking_dna:
83 out_fh.write( flanking_dna )
84
85 out_fh.close()
86
87 if __name__ == "__main__":
88 main()
89