Mercurial > repos > ieguinoa > data_manager_fetch_tx2gene
changeset 5:c380b7da5b65 draft default tip
Uploaded
| author | ieguinoa |
|---|---|
| date | Mon, 07 Jun 2021 16:33:53 +0000 |
| parents | bacd91d8b05a |
| children | |
| files | README.md data_manager/.Rhistory data_manager/data_manager_fetch_tx2gene.py data_manager_conf.xml tool_data_table_conf.xml.sample |
| diffstat | 4 files changed, 9 insertions(+), 12 deletions(-) [+] |
line wrap: on
line diff
--- a/README.md Sun Oct 21 17:48:48 2018 -0400 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,2 +0,0 @@ -# data_manager_fetch_tx2gene -Load entries in tx2gene table with transcript to gene tables or GTF/GFF file.
--- a/data_manager/data_manager_fetch_tx2gene.py Sun Oct 21 17:48:48 2018 -0400 +++ b/data_manager/data_manager_fetch_tx2gene.py Mon Jun 07 16:33:53 2021 +0000 @@ -27,7 +27,7 @@ CHUNK_SIZE = 2**20 # 1mb -DATA_TABLE_NAME = 'tx2gene' +DATA_TABLE_NAME = 'tx2gene_table' def cleanup_before_exit( tmp_dir ): if tmp_dir and os.path.exists( tmp_dir ): @@ -45,14 +45,14 @@ sequence_id = params['param_dict']['sequence_id'] if not sequence_id: sequence_id = dbkey #uuid.uuid4() generate and use an uuid instead? - + # if params['param_dict']['dbkey_source']['dbkey_source_selector'] == 'new': # dbkey_name = params['param_dict']['dbkey_source']['dbkey_name'] # if not dbkey_name: # dbkey_name = dbkey # else: # dbkey_name = None - dbkey = params['param_dict']['dbkey'] + dbkey = params['param_dict']['dbkey'] dbkey_name = dbkey_description sequence_name = params['param_dict']['sequence_name'] if not sequence_name: @@ -290,7 +290,6 @@ parser.add_option( '-b', '--base_dir', dest='base_dir', action='store', type='string', default=None, help='base_dir') parser.add_option( '-t', '--type', dest='file_type', action='store', type='string', default=None, help='file_type') (options, args) = parser.parse_args() - filename = args[0] #global DATA_TABLE_NAME rscript_gff_to_tx2gene=os.path.join( options.base_dir, 'get_tx2gene_table.R') @@ -298,14 +297,14 @@ #input_type='gff_gtf' #if options.file_type != 'gff_gtf': # file_type='tx2gene' - + params = loads( open( filename ).read() ) target_directory = params[ 'output_data' ][0]['extra_files_path'] os.mkdir( target_directory ) data_manager_dict = {} - + dbkey, dbkey_name, sequence_id, sequence_name = get_dbkey_dbname_id_name( params, dbkey_description=options.dbkey_description ) - + if dbkey in [ None, '', '?' ]: raise Exception( '"%s" is not a valid dbkey. You must specify a valid dbkey.' % ( dbkey ) ) @@ -318,6 +317,6 @@ cleanup_before_exit(tmp_dir) #save info to json file open( filename, 'wb' ).write( dumps( data_manager_dict ).encode() ) - + if __name__ == "__main__": main()
--- a/data_manager_conf.xml Sun Oct 21 17:48:48 2018 -0400 +++ b/data_manager_conf.xml Mon Jun 07 16:33:53 2021 +0000 @@ -1,7 +1,7 @@ <?xml version="1.0"?> <data_managers> <data_manager tool_file="data_manager/data_manager_fetch_tx2gene.xml" id="data_manager_fetch_tx2gene"> - <data_table name="tx2gene"> + <data_table name="tx2gene_table"> <output> <column name="value" /> <column name="dbkey" />
--- a/tool_data_table_conf.xml.sample Sun Oct 21 17:48:48 2018 -0400 +++ b/tool_data_table_conf.xml.sample Mon Jun 07 16:33:53 2021 +0000 @@ -1,4 +1,4 @@ <?xml version="1.0"?> <tables> - <table name="tx2gene" comment_char="#" allow_duplicate_entries="False"><columns>value, dbkey, name, path</columns><file path="tool-data/tx2gene.loc" /></table> + <table name="tx2gene_table" comment_char="#" allow_duplicate_entries="False"><columns>value, dbkey, name, path</columns><file path="tool-data/tx2gene.loc" /></table> </tables>