Mercurial > repos > dfornika > data_manager_build_kraken2_database
changeset 31:bb9833a41c6a draft
planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_build_kraken2_database/ commit 5629d9d7d9836d32b1e501277e213ef213de3e76-dirty
author | dfornika |
---|---|
date | Tue, 30 Apr 2019 14:16:12 -0400 |
parents | 0aaff8fb824d |
children | 23c68bbf573d |
files | data_manager/kraken2_build_database.py data_manager/kraken2_build_database.xml |
diffstat | 2 files changed, 164 insertions(+), 0 deletions(-) [+] |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/data_manager/kraken2_build_database.py Tue Apr 30 14:16:12 2019 -0400 @@ -0,0 +1,113 @@ +#!/usr/bin/env python + +from __future__ import print_function + +import argparse +import datetime +import errno +import json +import os +import subprocess + + +DATA_TABLE_NAME = "kraken2_databases" + + +def kraken2_build_standard(data_manager_dict, kraken2_args, target_directory, data_table_name=DATA_TABLE_NAME): + now = datetime.datetime.utcnow().strftime("%Y-%m-%dT%H%M%SZ") + + database_value = "_".join([ + now, + "standard", + "kmer-len" + str(kraken2_args["kmer_len"]), + "minimizer-len" + str(kraken2_args["minimizer_len"]), + "minimizer-spaces" + str(kraken2_args["minimizer_spaces"]), + ]) + + database_name = " ".join([ + "Standard", + "(Created:", + now + ",", + "kmer-len=" + str(kraken2_args["kmer_len"]) + ",", + "minimizer-len=" + str(kraken2_args["minimizer_len"]) + ",", + "minimizer-spaces=" + str(kraken2_args["minimizer_spaces"]) + ")", + ]) + + database_path = database_value + + args = [ + '--threads', str(kraken2_args["threads"]), + '--standard', + '--kmer-len', str(kraken2_args["kmer_len"]), + '--minimizer-len', str(kraken2_args["minimizer_len"]), + '--minimizer-spaces', str(kraken2_args["minimizer_spaces"]), + '--db', database_path + ] + + subprocess.check_call(['kraken2-build'] + args, cwd=target_directory) + + args = [ + '--threads', str(kraken2_args["threads"]), + '--clean', + '--db', database_path + ] + + subprocess.check_call(['kraken2-build'] + args, cwd=target_directory) + + data_table_entry = { + "value": database_value, + "name": database_name, + "path": database_path, + } + + _add_data_table_entry(data_manager_dict, data_table_entry) + + +def _add_data_table_entry(data_manager_dict, data_table_entry, data_table_name=DATA_TABLE_NAME): + 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 = argparse.ArgumentParser() + parser.add_argument('data_manager_json') + parser.add_argument( '-k', '--kmer-len', dest='kmer_len', type=int, default=35, help='kmer length' ) + parser.add_argument( '-m', '--minimizer-len', dest='minimizer_len', type=int, default=31, help='minimizer length' ) + parser.add_argument( '-s', '--minimizer-spaces', dest='minimizer_spaces', default=6, help='minimizer spaces' ) + parser.add_argument( '-t', '--threads', dest='threads', default=1, help='threads' ) + args = parser.parse_args() + + kraken2_args = { + "kmer_len": args.kmer_len, + "minimizer_len": args.minimizer_len, + "minimizer_spaces": args.minimizer_spaces, + "threads": args.threads, + } + + data_manager_input = json.loads(open(args.data_manager_json).read()) + + target_directory = data_manager_input['output_data'][0]['extra_files_path'] + + try: + os.mkdir( target_directory ) + except OSError as exc: + if exc.errno == errno.EEXIST and os.path.isdir( target_directory ): + pass + else: + raise + + data_manager_output = {} + + kraken2_build_standard( + data_manager_output, + kraken2_args, + target_directory, + ) + + open(args.data_manager_json, 'w').write(json.dumps(data_manager_output)) + + +if __name__ == "__main__": + main()
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/data_manager/kraken2_build_database.xml Tue Apr 30 14:16:12 2019 -0400 @@ -0,0 +1,51 @@ +<?xml version="1.0"?> +<tool id="kraken2_build_database" name="Kraken2 Database Builder" tool_type="manage_data" version="2.0.7_beta"> + <description>database builder</description> + <requirements> + <requirement type="package" version="2.0.7_beta">kraken2</requirement> + </requirements> + <version_command>kraken2 -version | head -n 1 | awk '{print $NF}'</version_command> + <command detect_errors="exit_code"> + <![CDATA[ + python '$__tool_directory__/kraken2_build_database.py' + '${out_file}' + --threads \${GALAXY_SLOTS:-1} + --kmer-len ${kmer_len} + --minimizer-len ${minimizer_len} + --minimizer-spaces ${minimizer_spaces} + ]]> + </command> + <inputs> + <conditional name="database_type"> + <param name="database_type" type="select" multiple="false" label="Database Type"> + <option value="standard">Standard</option> + <option value="special">Special</option> + <option value="custom">Custom</option> + </param> + <when value="standard"> + </when> + <when value="special"> + <param name="db_type" type="select" multiple="false" label="Select database to build"> + <option value="greengenes">Greengenes</option> + <option value="silva">Silva</option> + <option value="rdp">RDP</option> + </param> + </when> + <when value="custom"> + <param name="input_fasta" type="data" format="fasta" multiple="False" optional="true" label="Select history item" /> + <param name="database_name" type="text" label="Name for this database" /> + </when> + <param name="kmer_len" type="integer" value="35" label="K-mer length in BP" /> + <param name="minimizer_len" type="integer" value="31" label="Minimizer length" /> + <param name="minimizer_spaces" type="integer" value="6" label="Minimizer spaces" /> + </conditional> + </inputs> + <outputs> + <data name="out_file" format="data_manager_json" /> + </outputs> + <help> + </help> + <citations> + <citation type="doi">10.1186/gb-2014-15-3-r46</citation> + </citations> +</tool>