Mercurial > repos > iuc > data_manager_snpeff
view data_manager/data_manager_snpEff_databases.py @ 9:08d7998c3afb draft
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_snpeff commit 036feef4f46b5aedabc5080c8fa4bc00d9c30d95"
author | iuc |
---|---|
date | Wed, 12 Feb 2020 18:33:53 -0500 |
parents | 2107b4f40945 |
children | c6fbc5421697 |
line wrap: on
line source
#!/usr/bin/env python import json import optparse import os import subprocess import sys def fetch_databases(data_manager_dict, target_directory): if not os.path.exists(target_directory): os.makedirs(target_directory) databases_path = os.path.join(target_directory, 'databases.out') databases_output = open(databases_path, 'w') args = ['snpEff', 'databases'] return_code = subprocess.call(args=args, shell=False, stdout=databases_output.fileno()) if return_code: sys.exit(return_code) databases_output.close() data_manager_dict['data_tables'] = data_manager_dict.get('data_tables', {}) data_manager_dict['data_tables']['snpeffv_databases'] = data_manager_dict['data_tables'].get('snpeffv_databases', []) data_table_entries = [] with open(databases_path, 'r') as fh: for i, line in enumerate(fh): fields = line.split('\t') if len(fields) >= 2: genome_version = fields[0].strip() if genome_version.startswith("Genome") or genome_version.startswith("-"): continue # snpeff test genome if genome_version == '30c2c903' or fields[1].strip() == 'TestCase' or fields[1].strip().startswith('Test_'): continue description = fields[1].strip() + ' : ' + genome_version data_table_entries.append(dict(value=genome_version, name=description)) data_manager_dict['data_tables']['snpeffv_databases'] = data_table_entries return data_manager_dict def main(): parser = optparse.OptionParser() (options, args) = parser.parse_args() filename = args[0] params = json.loads(open(filename).read()) target_directory = params['output_data'][0]['extra_files_path'] os.mkdir(target_directory) data_manager_dict = {} # Create Defuse Reference Data data_manager_dict = fetch_databases(data_manager_dict, target_directory) # save info to json file open(filename, 'w').write(json.dumps(data_manager_dict, sort_keys=True)) if __name__ == "__main__": main()