Mercurial > repos > dave > data_manager_make_blastn_dbs
changeset 1:125b8ae43807 draft default tip
planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_fetch_genome_dbkeys_all_fasta commit db8f5e7fadc3bf752951b837a977919f8409e4b6-dirty
| author | dave |
|---|---|
| date | Sat, 30 Jun 2018 18:52:10 -0400 |
| parents | 3da261f8ac90 |
| children | |
| files | data_manager/.data_manager_make_blastn_dbs.py.swp data_manager/.data_manager_make_blastn_dbs.xml.swp data_manager/data_manager_make_blastn_database.py data_manager/data_manager_make_blastn_database.xml data_manager/data_manager_make_blastn_dbs.py data_manager/data_manager_make_blastn_dbs.xml data_manager_conf.xml tool_data_table_conf.xml.sample |
| diffstat | 8 files changed, 128 insertions(+), 141 deletions(-) [+] |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/data_manager/data_manager_make_blastn_database.py Sat Jun 30 18:52:10 2018 -0400 @@ -0,0 +1,65 @@ +#!/usr/bin/env python +# Dan Blankenberg +from __future__ import print_function + +import optparse +import os +import subprocess +import sys +from json import dumps, loads + +DEFAULT_DATA_TABLE_NAMES = ["blastn_databases"] + + +def build_blastn_index(data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_names=DEFAULT_DATA_TABLE_NAMES): + # TODO: allow multiple FASTA input files + fasta_base_name = os.path.split(fasta_filename)[-1] + sym_linked_fasta_filename = os.path.join(target_directory, fasta_base_name) + os.symlink(fasta_filename, sym_linked_fasta_filename) + args = ['makeblastdb', '-in', sym_linked_fasta_filename, '-dbtype', 'nucl', '-out', sequence_id] + proc = subprocess.Popen(args=args, shell=False, cwd=target_directory) + return_code = proc.wait() + if return_code: + print("Error building index.", file=sys.stderr) + sys.exit(return_code) + data_table_entry = dict(value=sequence_id, dbkey=dbkey, name=sequence_name, path=sequence_id) + for data_table_name in data_table_names: + _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry) + + +def _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry): + data_manager_dict['data_tables'] = data_manager_dict.get('data_tables', {}) + data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get(data_table_name, []) + data_manager_dict['data_tables'][data_table_name].append(data_table_entry) + return data_manager_dict + + +def main(): + parser = optparse.OptionParser() + parser.add_option('-i', '--input', dest='input', action='store', type="string", default=None, help='Input FASTA') + parser.add_option('-d', '--dbkey', dest='dbkey', action='store', type="string", default=None, help='Unique identifier') + parser.add_option('-n', '--name', dest='name', action='store', type="string", default=None, help='Name') + (options, args) = parser.parse_args() + + filename = args[0] + + params = loads(open(filename).read()) + #raise NameError(dumps(params, indent=2, sort_keys=True)) + target_directory = params['output_data'][0]['extra_files_path'] + os.mkdir(target_directory) + data_manager_dict = {} + + if options.dbkey in [None, '', '?']: + raise Exception('"%s" is not a valid dbkey. You must specify a valid dbkey.' % (options.dbkey)) + + sequence_id, sequence_name = (options.dbkey, options.name) + + # build the index + build_blastn_index(data_manager_dict, options.input, params, target_directory, options.dbkey, sequence_id, sequence_name, data_table_names=['blastn_databases']) + + # save info to json file + open(filename, 'wb').write(dumps(data_manager_dict)) + + +if __name__ == "__main__": + main()
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/data_manager/data_manager_make_blastn_database.xml Sat Jun 30 18:52:10 2018 -0400 @@ -0,0 +1,61 @@ +<tool id="data_manager_make_blastn_database" name="Create blastn database" version="0.0.1" tool_type="manage_data"> + <description>from fasta file</description> + <requirements> + <requirement type="package" version="2.6.0">blast</requirement> + </requirements> + <command detect_errors="exit_code"><![CDATA[ + python '$__tool_directory__/data_manager_make_blastn_database.py' + '${out_file}' + #if str( $input_source.source_selector ) == 'history': + --input '$input_source.input' + --dbkey '$input_source.dbkey' + --name '$input_source.name' + #else + --input '${all_fasta_source.fields.path}' + --dbkey '${all_fasta_source.fields.dbkey}' + --name '${all_fasta_source.fields.name}' + #end if + ]]></command> + <inputs> + <conditional name="input_source"> + <param name="source_selector" type="select" label="Use locally cached fasta or history dataset"> + <option value="history" selected="True">From history</option> + <option value="data_table">From data table</option> + </param> + <when value="data_table"> + <param name="input" type="data" label="Cached sequence"> + <options from_data_table="all_fasta" /> + </param> + </when> + <when value="history"> + <param name="input" type="data" format="fasta" optional="false" /> + <param name="dbkey" type="text" label="Unique identifier" /> + <param name="name" type="text" label="Display name for database" /> + </when> + </conditional> + </inputs> + <outputs> + <data name="out_file" format="data_manager_json"/> + </outputs> + <tests> + <!-- TODO: need some way to test that new entry was added to data table --> + <test> + <param name="source_selector" value="history" /> + <param name="input" value="phiX174.fasta" ftype="fasta" /> + <param name="dbkey" value="phiX174"/> + <param name="name" value="Phi-X 174 12-06-1912"/> + <output name="out_file" file="phiX174.data_manager_json"/> + </test> + </tests> + <help> +**What it does** + +Fetches a reference genomes or sequences from various sources (URL, Galaxy History, or a server directory) and populates the "blastn_db" data table. + +------ + +.. class:: infomark + +**Notice:** If you leave name, description, or id blank, it will be generated automatically. + </help> +</tool>
--- a/data_manager/data_manager_make_blastn_dbs.py Sat Jun 30 16:19:53 2018 -0400 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,78 +0,0 @@ -#!/usr/bin/env python -# Dan Blankenberg -from __future__ import print_function - -import optparse -import os -import subprocess -import sys -from json import dumps, loads - -DEFAULT_DATA_TABLE_NAMES = ["blastn_dbs"] - - -def get_id_name(params, dbkey, fasta_description=None): - # TODO: ensure sequence_id is unique and does not already appear in location file - sequence_id = params['param_dict']['sequence_id'] - if not sequence_id: - sequence_id = dbkey - - sequence_name = params['param_dict']['sequence_name'] - if not sequence_name: - sequence_name = fasta_description - if not sequence_name: - sequence_name = dbkey - return sequence_id, sequence_name - - -def build_blastn_index(data_manager_dict, fasta_filename, params, target_directory, dbkey, sequence_id, sequence_name, data_table_names=DEFAULT_DATA_TABLE_NAMES): - # TODO: allow multiple FASTA input files - fasta_base_name = os.path.split(fasta_filename)[-1] - sym_linked_fasta_filename = os.path.join(target_directory, fasta_base_name) - os.symlink(fasta_filename, sym_linked_fasta_filename) - args = ['makeblastdb', '-in', sym_linked_fasta_filename, '-dbtype', "'nucl'", '-out', sequence_id] - proc = subprocess.Popen(args=args, shell=False, cwd=target_directory) - return_code = proc.wait() - if return_code: - print("Error building index.", file=sys.stderr) - sys.exit(return_code) - data_table_entry = dict(value=sequence_id, dbkey=dbkey, name=sequence_name, path=sequence_id) - for data_table_name in data_table_names: - _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry) - - -def _add_data_table_entry(data_manager_dict, data_table_name, data_table_entry): - data_manager_dict['data_tables'] = data_manager_dict.get('data_tables', {}) - data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get(data_table_name, []) - data_manager_dict['data_tables'][data_table_name].append(data_table_entry) - return data_manager_dict - - -def main(): - parser = optparse.OptionParser() - parser.add_option('-i', '--input', dest='input', action='store', type="string", default=None, help='Input FASTA') - parser.add_option('-d', '--dbkey', dest='dbkey', action='store', type="string", default=None, help='Unique identifier') - parser.add_option('-n', '--name', dest='name', action='store', type="string", default=None, help='Name') - (options, args) = parser.parse_args() - - filename = args[0] - - params = loads(open(filename).read()) - target_directory = params['output_data'][0]['extra_files_path'] - os.mkdir(target_directory) - data_manager_dict = {} - - if dbkey in [None, '', '?']: - raise Exception('"%s" is not a valid dbkey. You must specify a valid dbkey.' % (dbkey)) - - sequence_id, sequence_name = get_id_name(params, dbkey=options.dbkey, fasta_description=options.name) - - # build the index - build_blastn_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) - - # save info to json file - open(filename, 'wb').write(dumps(data_manager_dict)) - - -if __name__ == "__main__": - main()
--- a/data_manager/data_manager_make_blastn_dbs.xml Sat Jun 30 16:19:53 2018 -0400 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,61 +0,0 @@ -<tool id="data_manager_make_blastn_dbs" name="Create blastn database" version="0.0.1" tool_type="manage_data"> - <description>for fasta and blastdbs</description> - <requirements> - <requirement type="package" version="2.6.0">blast</requirement> - </requirements> - <command detect_errors="exit_code"><![CDATA[ - python '$__tool_directory__/data_manager_make_blastn_dbs.py' - '${out_file}' - #if str( $input_source.source_selector ) == 'history': - --input '$input_source.input' - --dbkey '$input_source.dbkey' - --name '$input_source.name' - #else - --input '${all_fasta_source.fields.path}' - --dbkey '${all_fasta_source.fields.dbkey}' - --name '${all_fasta_source.fields.name}' - #end if - ]]></command> - <inputs> - <conditional name="input_source"> - <param name="source_selector" type="select" label="Use locally cached fasta or history dataset"> - <option value="history" selected="True">From history</option> - <option value="data_table">From data table</option> - </param> - <when value="data_table"> - <param name="input" type="data" label="Cached sequence"> - <options from_data_table="all_fasta" /> - </param> - </when> - <when value="history"> - <param name="input" type="data" format="fasta" optional="false" /> - <param name="dbkey" type="text" label="Unique identifier" /> - <param name="name" type="text" label="Display name for database" /> - </when> - </conditional> - </inputs> - <outputs> - <data name="out_file" format="data_manager_json"/> - </outputs> - <tests> - <!-- TODO: need some way to test that new entry was added to data table --> - <test> - <param name="source_selector" value="history" /> - <param name="input" value="phiX174.fasta" ftype="fasta" /> - <param name="dbkey" value="phiX174"/> - <param name="name" value="Phi-X 174 12-06-1912"/> - <output name="out_file" file="phiX174.data_manager_json"/> - </test> - </tests> - <help> -**What it does** - -Fetches a reference genomes or sequences from various sources (URL, Galaxy History, or a server directory) and populates the "blastn_db" data table. - ------- - -.. class:: infomark - -**Notice:** If you leave name, description, or id blank, it will be generated automatically. - </help> -</tool>
--- a/data_manager_conf.xml Sat Jun 30 16:19:53 2018 -0400 +++ b/data_manager_conf.xml Sat Jun 30 18:52:10 2018 -0400 @@ -1,6 +1,6 @@ <?xml version="1.0"?> <data_managers> - <data_manager tool_file="data_manager/data_manager_make_blastn_dbs.xml" id="data_manager_make_blastn_dbs"> + <data_manager tool_file="data_manager/data_manager_make_blastn_database.xml" id="data_manager_make_blastn_database"> <data_table name="all_fasta"> <output> <column name="value" />
--- a/tool_data_table_conf.xml.sample Sat Jun 30 16:19:53 2018 -0400 +++ b/tool_data_table_conf.xml.sample Sat Jun 30 18:52:10 2018 -0400 @@ -7,6 +7,6 @@ <!-- Locations of dbkeys and len files under genome directory --> <table name="blastn_databases" comment_char="#"> <columns>value, name, path</columns> - <file path="tool-data/blastn_dbs.loc" /> + <file path="tool-data/blastn_databases.loc" /> </table> </tables>