main.py

#!/usr/bin/env python3

import argparse
import configparser
import sys
import os
import platform
import numpy as np


# modules
from PCprophet import io_ as io
from PCprophet import collapse as collapse
from PCprophet import generate_features_v2 as generate_features
from PCprophet import hypothesis as hypothesis
from PCprophet import map_to_database as map_to_database
from PCprophet import merge as merge
from PCprophet import differential as differential
from PCprophet import predict as predict
from PCprophet import plots as plots

from PCprophet import validate_input as validate


class ParserHelper(argparse.ArgumentParser):
    def error(self, message):
        sys.stderr.write('error: %s\n' % message)
        self.print_help()
        sys.exit(2)


# TODO check os
def get_os():
    return platform.system()


def create_config():
    '''
    parse command line and create .ini file for configuration
    '''
    parser = ParserHelper(description='Protein Complex Prophet argument')
    parser.add_argument(
        '-db',
        help='protein complex database from CORUM or ppi network in STRING format',
        dest='database',
        action='store',
        default='coreComplexes.txt',
    )
    # maybe better to add function for generating a dummy sample id?
    parser.add_argument(
        '-sid',
        help='sample ids file',
        dest='sample_ids',
        default='sample_ids.txt',
        action='store',
    )
    parser.add_argument(
        '-output',
        help='outfile folder path',
        dest='out_folder',
        default=r'./Output',
        action='store',
    )
    # TODO change tmp to Output/tmp check resource_path important for windows
    parser.add_argument(
        '-cal',
        help='calibration file no headers tab delimited fractiosn to mw in KDa',
        dest='calibration',
        default='None',
        action='store',
    )
    parser.add_argument(
        '-mw_uniprot',
        help='Molecular weight from uniprot',
        dest='mwuni',
        default='None',
        action='store',
    )
    parser.add_argument(
        '-is_ppi',
        help='is the -db a protein protein interaction database',
        dest='is_ppi',
        action='store',
        default='False',
        choices=['True', 'False'],
    )
    parser.add_argument(
        '-a',
        help='use all fractions [1,X]',
        dest='all_fract',
        action='store',
        default='all',
    )
    parser.add_argument(
        '-ma',
        help='merge using all complexes or reference only',
        dest='merge',
        action='store',
        choices=['all', 'reference'],
        default='all',
    )
    parser.add_argument(
        '-fdr',
        help='false discovery rate for novel complexes',
        dest='fdr',
        action='store',
        default=0.5,
        type=float,
    )
    parser.add_argument(
        '-co',
        help='collapse mode',
        choices=['GO', 'CAL', 'SUPER', 'PROB', 'NONE'],
        dest='collapse',
        default='GO',
        action='store',
    )
    parser.add_argument(
        '-sc',
        help='score for missing proteins in differential analysis',
        dest='score_missing',
        action='store',
        default=0.5,
        type=float,
    )
    parser.add_argument(
        '-mult',
        help='Multi processing feature generation',
        dest='multi',
        action='store',
        default='True',
        choices=['True', 'False'],
    )
    parser.add_argument('-w', dest='weight_pred', help='LEGACY', action='store', default=1, type=float)
    parser.add_argument('-v', dest='verbose', help='Verbose', action='store', default=1)
    parser.add_argument('-skip',
                        dest='skip',
                        help='Skip feature generation and complex prediction step',action='store',
                        default=False)
    args = parser.parse_args()

    # deal with numpy warnings and so on
    if args.verbose == 0:
        np.seterr(all='ignore')
    else:
        pass
        # print them

    # create config file
    config = configparser.ConfigParser()
    config['GLOBAL'] = {
        'db': args.database,
        'sid': args.sample_ids,
        'go_obo': io.resource_path('go-basic.obo'),
        'sp_go': io.resource_path('tmp_GO_sp_only.txt'),
        'output': args.out_folder,
        'cal': args.calibration,
        'mw': args.mwuni,
        'temp': r'./tmp',
        'mult': args.multi,
        'skip': args.skip
    }
    config['PREPROCESS'] = {
        'is_ppi': args.is_ppi,
        'all_fract': args.all_fract,
        'merge': args.merge,
    }
    config['POSTPROCESS'] = {'fdr': args.fdr, 'collapse_mode': args.collapse}
    config['DIFFERENTIAL'] = {
        'score_missing': args.score_missing,
        'weight_pred': args.weight_pred,
        'fold_change': '-5,-2,2,5',
        'correlation': '0.3,0.9',
        'ratio': '-2,-0.5,0.5,2',
        'shift': '-10,-5,5,10',
        'weight_fold_change': 1,
        'weight_correlation': 0.75,
        'weight_ratio': 0.25,
        'weight_shift': 0.5,
    }
    # create config ini file for backup
    with open('ProphetConfig.conf', 'w') as conf:
        config.write(conf)
    return config


def preprocessing(infile, config):
    #validate.InputTester(infile, 'in').test_file()
    map_to_database.runner(
        infile=infile,
        db=config['GLOBAL']['db'],
        is_ppi=config['PREPROCESS']['is_ppi'],
        use_fr=config['PREPROCESS']['all_fract'],
    )
    hypothesis.runner(
        infile=infile,
        hypothesis=config['PREPROCESS']['merge'],
        use_fr=config['PREPROCESS']['all_fract'],
    )
    #  # sample specific folder
    tmp_folder = io.file2folder(infile, prefix=config['GLOBAL']['temp'])
    merge.runner(base=tmp_folder, mergemode=config['PREPROCESS']['merge'])
    generate_features.runner(
        tmp_folder,
        config['GLOBAL']['go_obo'],
        config['GLOBAL']['sp_go'],
        config['GLOBAL']['mult']
    )
    predict.runner(tmp_folder)
    return True


def main():
    config = create_config()
    validate.InputTester(config['GLOBAL']['db'], 'db').test_file()
    validate.InputTester(config['GLOBAL']['sid'], 'ids').test_file()
    files = io.read_sample_ids(config['GLOBAL']['sid'])
    files = [os.path.abspath(x) for x in files.keys()]
    # skip feature generation
    if config['GLOBAL']['skip'] == 'False':
        [preprocessing(infile, config) for infile in files]
    collapse.runner(
        config['GLOBAL']['temp'],
        config['GLOBAL']['sid'],
        config['GLOBAL']['cal'],
        config['GLOBAL']['mw'],
        config['POSTPROCESS']['fdr'],
        config['POSTPROCESS']['collapse_mode'],
    )
    combined_file = os.path.join(config['GLOBAL']['temp'], 'combined.txt')
    differential.runner(
        combined_file,
        config['GLOBAL']['sid'],
        config['GLOBAL']['output'],
        config['GLOBAL']['temp'],
    )
    plots.runner(
        config['GLOBAL']['temp'],
        config['GLOBAL']['output'],
        config['POSTPROCESS']['fdr'],
        config['GLOBAL']['sid'],
    )


if __name__ == '__main__':
    try:
        main()
    except KeyboardInterrupt:
        pass