Mercurial > repos > matthias > data_manager_bowtie2_index_builder
comparison data_manager/bowtie2_index_builder.py @ 0:68fd790f3916 draft default tip
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_bowtie2_index_builder commit eb0405e83bae7b6f11446236044a6aed2fcaf4d6"
author | matthias |
---|---|
date | Thu, 02 Apr 2020 13:16:53 +0000 |
parents | |
children |
comparison
equal
deleted
inserted
replaced
-1:000000000000 | 0:68fd790f3916 |
---|---|
1 #!/usr/bin/env python | |
2 # Dan Blankenberg | |
3 from __future__ import print_function | |
4 | |
5 import optparse | |
6 import os | |
7 import subprocess | |
8 import sys | |
9 from json import dumps, loads | |
10 | |
11 DEFAULT_DATA_TABLE_NAMES = ["bowtie2_indexes"] | |
12 | |
13 | |
14 def get_id_name(params, dbkey, fasta_description=None): | |
15 # TODO: ensure sequence_id is unique and does not already appear in location file | |
16 sequence_id = params['param_dict']['sequence_id'] | |
17 if not sequence_id: | |
18 sequence_id = dbkey | |
19 | |
20 sequence_name = params['param_dict']['sequence_name'] | |
21 if not sequence_name: | |
22 sequence_name = fasta_description | |
23 if not sequence_name: | |
24 sequence_name = dbkey | |
25 return sequence_id, sequence_name | |
26 | |
27 | |
28 def build_bowtie2_index(data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_names=DEFAULT_DATA_TABLE_NAMES): | |
29 # TODO: allow multiple FASTA input files | |
30 fasta_base_name = os.path.split(fasta_filename)[-1] | |
31 sym_linked_fasta_filename = os.path.join(target_directory, fasta_base_name) | |
32 os.symlink(fasta_filename, sym_linked_fasta_filename) | |
33 args = ['bowtie2-build', sym_linked_fasta_filename, sequence_id] | |
34 threads = os.environ.get('GALAXY_SLOTS') | |
35 if threads: | |
36 args.extend(['--threads', threads]) | |
37 proc = subprocess.Popen(args=args, shell=False, cwd=target_directory) | |
38 return_code = proc.wait() | |
39 if return_code: | |
40 print("Error building index.", file=sys.stderr) | |
41 sys.exit(return_code) | |
42 data_table_entry = dict(value=sequence_id, dbkey=dbkey, name=sequence_name, path=sequence_id) | |
43 for data_table_name in data_table_names: | |
44 _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry) | |
45 | |
46 | |
47 def _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry): | |
48 data_manager_dict['data_tables'] = data_manager_dict.get('data_tables', {}) | |
49 data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get(data_table_name, []) | |
50 data_manager_dict['data_tables'][data_table_name].append(data_table_entry) | |
51 return data_manager_dict | |
52 | |
53 | |
54 def main(): | |
55 parser = optparse.OptionParser() | |
56 parser.add_option('-f', '--fasta_filename', dest='fasta_filename', action='store', type="string", default=None, help='fasta_filename') | |
57 parser.add_option('-d', '--fasta_dbkey', dest='fasta_dbkey', action='store', type="string", default=None, help='fasta_dbkey') | |
58 parser.add_option('-t', '--fasta_description', dest='fasta_description', action='store', type="string", default=None, help='fasta_description') | |
59 parser.add_option('-n', '--data_table_name', dest='data_table_name', action='append', type="string", default=None, help='data_table_name') | |
60 (options, args) = parser.parse_args() | |
61 | |
62 filename = args[0] | |
63 | |
64 params = loads(open(filename).read()) | |
65 target_directory = params['output_data'][0]['extra_files_path'] | |
66 os.mkdir(target_directory) | |
67 data_manager_dict = {} | |
68 | |
69 dbkey = options.fasta_dbkey | |
70 | |
71 if dbkey in [None, '', '?']: | |
72 raise Exception('"%s" is not a valid dbkey. You must specify a valid dbkey.' % (dbkey)) | |
73 | |
74 sequence_id, sequence_name = get_id_name(params, dbkey=dbkey, fasta_description=options.fasta_description) | |
75 | |
76 # build the index | |
77 build_bowtie2_index(data_manager_dict, options.fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_names=options.data_table_name or DEFAULT_DATA_TABLE_NAMES) | |
78 | |
79 # save info to json file | |
80 with open(filename, 'w') as json_out: | |
81 json_out.write(dumps(data_manager_dict, sort_keys=True)) | |
82 | |
83 | |
84 if __name__ == "__main__": | |
85 main() |