Mercurial > repos > iuc > data_manager_snpeff

comparison data_manager/data_manager_snpEff_databases.py @ 0:111fc9457ec6 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/data_managers/data_manager_snpeff commit 88c982c5adcd32b11d98428fc554a4fdfcc19584
author iuc
date Tue, 07 Jun 2016 10:13:35 -0400
parents
children 821c34cdacb6
comparison
equal deleted inserted replaced
-1:000000000000 0:111fc9457ec6
1 #!/usr/bin/env python
2 import json
3 import optparse
4 import os
5 import subprocess
6 import sys
7
8
9 def stop_err(msg):
10 sys.stderr.write(msg)
11 sys.exit(1)
12
13
14 def fetch_databases(data_manager_dict, target_directory, jar_path):
15 (snpEff_dir, snpEff_jar) = os.path.split(jar_path)
16 if not os.path.exists(target_directory):
17 os.makedirs(target_directory)
18 databases_path = os.path.join( target_directory, 'databases.out' )
19 databases_output = open(databases_path, 'w')
20 args = [ 'java', '-jar' ]
21 args.append( snpEff_jar )
22 args.append( 'databases' )
23 # tmp_stderr = tempfile.NamedTemporaryFile( prefix = "tmp-data-manager-snpEff-stderr" )
24 # databases_output = open(databases_path)
25 # proc = subprocess.Popen( args=args, shell=False, cwd=snpEff_dir, stdout=databases_output.fileno(), stderr=tmp_stderr.fileno() )
26 proc = subprocess.Popen( args=args, shell=False, cwd=snpEff_dir, stdout=databases_output.fileno() )
27 return_code = proc.wait()
28 if return_code:
29 sys.exit( return_code )
30 databases_output.close()
31 data_manager_dict['data_tables'] = data_manager_dict.get( 'data_tables', {} )
32 data_manager_dict['data_tables']['snpeff4_databases'] = data_manager_dict['data_tables'].get( 'snpeff4_databases', [] )
33 data_table_entries = []
34 try:
35 fh = open(databases_path, 'r')
36 for i, line in enumerate(fh):
37 fields = line.split('\t')
38 if len(fields) >= 2:
39 genome_version = fields[0].strip()
40 if genome_version.startswith("Genome") or genome_version.startswith("-"):
41 continue
42 # snpeff test genome
43 if genome_version == '30c2c903' or fields[1].strip() == 'TestCase' or fields[1].strip().startswith('Test_'):
44 continue
45 description = fields[1].strip() + ' : ' + genome_version
46 data_table_entries.append(dict(value=genome_version, name=description))
47 data_manager_dict['data_tables']['snpeff4_databases'] = data_table_entries
48 except Exception as e:
49 stop_err( 'Error parsing %s %s\n' % (databases_path, str( e )) )
50 else:
51 fh.close()
52 return data_manager_dict
53
54
55 def main():
56 # Parse Command Line
57 parser = optparse.OptionParser()
58 parser.add_option( '-j', '--jar_path', dest='jar_path', action='store', type="string", default=None, help='snpEff.jar path' )
59 (options, args) = parser.parse_args()
60
61 filename = args[0]
62
63 params = json.loads( open( filename ).read() )
64 target_directory = params[ 'output_data' ][0]['extra_files_path']
65 os.mkdir( target_directory )
66 data_manager_dict = {}
67
68 # Create Defuse Reference Data
69 data_manager_dict = fetch_databases( data_manager_dict, target_directory, options.jar_path)
70
71 # save info to json file
72 open( filename, 'wb' ).write( json.dumps( data_manager_dict ) )
73
74 if __name__ == "__main__":
75 main()