Mercurial > repos > jjohnson > data_manager_picard_index_builder

comparison data_manager/picard_index_builder.py @ 0:ce8e6a5b8830 draft default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author jjohnson
date Tue, 20 Aug 2013 11:12:02 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:ce8e6a5b8830
1 #!/usr/bin/env python
2 #Jim Johnson
3
4 import sys
5 import os
6 import tempfile
7 import optparse
8 import subprocess
9
10 from galaxy.util.json import from_json_string, to_json_string
11
12 CHUNK_SIZE = 2**20
13
14 DEFAULT_DATA_TABLE_NAME = "picard_indexes"
15
16 def get_id_name( params, dbkey, fasta_description=None):
17 #TODO: ensure sequence_id is unique and does not already appear in location file
18 sequence_id = params['param_dict']['sequence_id']
19 if not sequence_id:
20 sequence_id = dbkey
21
22 sequence_name = params['param_dict']['sequence_name']
23 if not sequence_name:
24 sequence_name = fasta_description
25 if not sequence_name:
26 sequence_name = dbkey
27 return sequence_id, sequence_name
28
29 def build_picard_index( jar_path, data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_name=DEFAULT_DATA_TABLE_NAME):
30 #TODO: allow multiple FASTA input files
31 #tmp_dir = tempfile.mkdtemp( prefix='tmp-data-manager-picard-index-builder-' )
32 fasta_base_name = os.path.split( fasta_filename )[-1]
33 sym_linked_fasta_filename = os.path.join( target_directory, fasta_base_name )
34 os.symlink( fasta_filename, sym_linked_fasta_filename )
35 basename,ext = os.path.splitext(fasta_base_name)
36 ref_dict = "%s.dict" % basename
37 args = [ 'java','-Xmx4G','-jar', jar_path, "R=%s" % sym_linked_fasta_filename, "O=%s" % ref_dict ]
38 tmp_stderr = tempfile.NamedTemporaryFile( prefix = "tmp-data-manager-picard-index-builder-stderr" )
39 proc = subprocess.Popen( args=args, shell=False, cwd=target_directory, stderr=tmp_stderr.fileno() )
40 return_code = proc.wait()
41 if return_code:
42 tmp_stderr.flush()
43 tmp_stderr.seek(0)
44 print >> sys.stderr, "Error building index:"
45 while True:
46 chunk = tmp_stderr.read( CHUNK_SIZE )
47 if not chunk:
48 break
49 sys.stderr.write( chunk )
50 sys.exit( return_code )
51 tmp_stderr.close()
52 data_table_entry = dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_name )
53 _add_data_table_entry( data_manager_dict, data_table_name, data_table_entry )
54
55 def _add_data_table_entry( data_manager_dict, data_table_name, data_table_entry ):
56 data_manager_dict['data_tables'] = data_manager_dict.get( 'data_tables', {} )
57 data_manager_dict['data_tables'][ data_table_name ] = data_manager_dict['data_tables'].get( data_table_name, [] )
58 data_manager_dict['data_tables'][ data_table_name ].append( data_table_entry )
59 return data_manager_dict
60
61 def main():
62 #Parse Command Line
63 parser = optparse.OptionParser()
64 parser.add_option( '-f', '--fasta_filename', dest='fasta_filename', action='store', type="string", default=None, help='fasta_filename' )
65 parser.add_option( '-d', '--fasta_dbkey', dest='fasta_dbkey', action='store', type="string", default=None, help='fasta_dbkey' )
66 parser.add_option( '-t', '--fasta_description', dest='fasta_description', action='store', type="string", default=None, help='fasta_description' )
67 parser.add_option( '-n', '--data_table_name', dest='data_table_name', action='store', type="string", default=None, help='data_table_name' )
68 parser.add_option( '-j', '--jar_path', dest='jar_path', action='store', type="string", default=None, help='path to CreateSequenceDictionary.jar' )
69 (options, args) = parser.parse_args()
70
71 filename = args[0]
72
73 params = from_json_string( open( filename ).read() )
74 target_directory = params[ 'output_data' ][0]['extra_files_path']
75 os.mkdir( target_directory )
76 data_manager_dict = {}
77
78 dbkey = options.fasta_dbkey
79
80 if dbkey in [ None, '', '?' ]:
81 raise Exception( '"%s" is not a valid dbkey. You must specify a valid dbkey.' % ( dbkey ) )
82
83 sequence_id, sequence_name = get_id_name( params, dbkey=dbkey, fasta_description=options.fasta_description )
84
85 #build the index
86 build_picard_index( options.jar_path, data_manager_dict, options.fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_name=options.data_table_name or DEFAULT_DATA_TABLE_NAME )
87
88 #save info to json file
89 open( filename, 'wb' ).write( to_json_string( data_manager_dict ) )
90
91 if __name__ == "__main__": main()