Mercurial > repos > devteam > samtools_mpileup

comparison samtools_wrapper.py @ 0:44a18a94d7a9

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded tool tarball.
author devteam
date Mon, 26 Aug 2013 14:23:36 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:44a18a94d7a9
1 #!/usr/bin/env python
2 #Dan Blankenberg
3
4 """
5 A wrapper script for running SAMTools commands.
6 """
7
8 import sys, optparse, os, tempfile, subprocess, shutil
9 from string import Template
10
11 GALAXY_EXT_TO_SAMTOOLS_EXT = { 'bam_index':'bam.bai', } #items not listed here will use the galaxy extension as-is
12 GALAXY_EXT_TO_SAMTOOLS_FILE_TYPE = GALAXY_EXT_TO_SAMTOOLS_EXT #for now, these are the same, but could be different if needed
13 DEFAULT_SAMTOOLS_PREFIX = "SAMTools_file"
14 CHUNK_SIZE = 2**20 #1mb
15
16
17 def cleanup_before_exit( tmp_dir ):
18 if tmp_dir and os.path.exists( tmp_dir ):
19 shutil.rmtree( tmp_dir )
20
21 def SAMTOOLS_filename_from_galaxy( galaxy_filename, galaxy_ext, target_dir = None, prefix = None ):
22 suffix = GALAXY_EXT_TO_SAMTOOLS_EXT.get( galaxy_ext, galaxy_ext )
23 if prefix is None:
24 prefix = DEFAULT_SAMTOOLS_PREFIX
25 if target_dir is None:
26 target_dir = os.getcwd()
27 SAMTools_filename = os.path.join( target_dir, "%s.%s" % ( prefix, suffix ) )
28 os.symlink( galaxy_filename, SAMTools_filename )
29 return SAMTools_filename
30
31 def SAMTOOLS_filetype_argument_substitution( argument, galaxy_ext ):
32 return argument % dict( file_type = GALAXY_EXT_TO_SAMTOOLS_FILE_TYPE.get( galaxy_ext, galaxy_ext ) )
33
34 def open_file_from_option( filename, mode = 'rb' ):
35 if filename:
36 return open( filename, mode = mode )
37 return None
38
39 def html_report_from_directory( html_out, dir ):
40 html_out.write( '<html>\n<head>\n<title>Galaxy - SAMTOOLS Output</title>\n</head>\n<body>\n<p/>\n<ul>\n' )
41 for fname in sorted( os.listdir( dir ) ):
42 html_out.write( '<li><a href="%s">%s</a></li>\n' % ( fname, fname ) )
43 html_out.write( '</ul>\n</body>\n</html>\n' )
44
45 def __main__():
46 #Parse Command Line
47 parser = optparse.OptionParser()
48 parser.add_option( '-p', '--pass_through', dest='pass_through_options', action='append', type="string", help='These options are passed through directly to SAMTOOLS, without any modification.' )
49 parser.add_option( '-d', '--dataset', dest='datasets', action='append', type="string", nargs=4, help='"-argument" "original_filename" "galaxy_filetype" "name_prefix"' )
50 parser.add_option( '', '--stdout', dest='stdout', action='store', type="string", default=None, help='If specified, the output of stdout will be written to this file.' )
51 parser.add_option( '', '--stderr', dest='stderr', action='store', type="string", default=None, help='If specified, the output of stderr will be written to this file.' )
52 parser.add_option( '', '--html_report_from_directory', dest='html_report_from_directory', action='append', type="string", nargs=2, help='"Target HTML File" "Directory"')
53 (options, args) = parser.parse_args()
54
55 tmp_dir = tempfile.mkdtemp( prefix='tmp-SAMTOOLS-' )
56
57 #set up stdout and stderr output options
58 stdout = open_file_from_option( options.stdout, mode = 'wb' )
59 stderr = open_file_from_option( options.stderr, mode = 'wb' )
60 #if no stderr file is specified, we'll use our own
61 if stderr is None:
62 stderr = tempfile.NamedTemporaryFile( prefix="SAMTOOLS-stderr-", dir=tmp_dir )
63
64 if options.pass_through_options:
65 cmd = ' '.join( options.pass_through_options )
66 else:
67 cmd = ''
68 return_code = None
69 if options.datasets:
70 for ( dataset_arg, filename, galaxy_ext, prefix ) in options.datasets:
71 SAMTools_filename = SAMTOOLS_filename_from_galaxy( filename, galaxy_ext, target_dir = tmp_dir, prefix = prefix )
72 if dataset_arg:
73 if '>' in cmd:
74 cmd = cmd.replace( '>', ' %s "%s" >' % ( SAMTOOLS_filetype_argument_substitution( dataset_arg, galaxy_ext ), SAMTools_filename ), 1 )
75 else:
76 cmd = '%s %s "%s"' % ( cmd, SAMTOOLS_filetype_argument_substitution( dataset_arg, galaxy_ext ), SAMTools_filename )
77 #auto index fasta files:
78 if galaxy_ext == 'fa':
79 index_cmd = 'samtools faidx %s' % ( SAMTools_filename )
80 proc = subprocess.Popen( args=index_cmd, stdout=stdout, stderr=stderr, shell=True, cwd=tmp_dir )
81 return_code = proc.wait()
82 if return_code:
83 break
84 if return_code is None or not return_code:
85 proc = subprocess.Popen( args=cmd, stdout=stdout, stderr=stderr, shell=True, cwd=tmp_dir )
86 return_code = proc.wait()
87 if return_code:
88 stderr_target = sys.stderr
89 else:
90 if stdout:
91 stderr_target = stdout
92 else:
93 stderr_target = sys.stdout
94 stderr.flush()
95 stderr.seek(0)
96 while True:
97 chunk = stderr.read( CHUNK_SIZE )
98 if chunk:
99 stderr_target.write( chunk )
100 else:
101 break
102 stderr.close()
103 #generate html reports
104 if options.html_report_from_directory:
105 for ( html_filename, html_dir ) in options.html_report_from_directory:
106 html_report_from_directory( open( html_filename, 'wb' ), html_dir )
107
108 cleanup_before_exit( tmp_dir )
109
110 if __name__=="__main__": __main__()