main.py

#!/usr/bin/env python3

#
# BRAID: A python program to find protein complexes from high-throughput data
# 
# Input file: a CSV file containing a list of bait-preys experiments
# The first protein at the beginning of every line is a bait protein.
#   
# Model parameters:
# 1. False negative rate
# 2. False positive rate
# 3. A maximum possible number of clusters

import argparse
import numpy as np
import matplotlib.pyplot as plt
from scipy import stats

import braidmrf.observation.countSpokeModel as cpm
import braidmrf.observation.countBioplexSpokeModel as cpmBioplex
import braidmrf.meanfield.simulateLikelihood as smlt

import braidmrf.inputFile.inputFile as inputFile
import braidmrf.inputFile.inputBioplex as inputBioplex

from time import time as timer

def clustering(inputSet, Nk, psi):
    fn = 0.8
    fp = 0.04
    nProteins = inputSet.observationG.nProteins
    cmfa = smlt.CMeanFieldAnnealing(nProteins, Nk)
    ts = timer()
    cmfa.estimate(inputSet.observationG, nProteins, Nk, psi)
    te = timer()
    print("Time running MFA: ", te-ts)

    (fn, fp) = cmfa.computeResidues(inputSet.observationG, nProteins, Nk)
    cmfa.computeEntropy(nProteins, Nk)
    
    print("False negative rate = " + str(fn))
    print("False positive rate = " + str(fp))
    
    inputSet.writeCluster2File(cmfa.mIndicatorQ, cmfa.indicatorVec)
    inputSet.observationG.write2cytoscape(cmfa.indicatorVec, cmfa.mIndicatorQ, inputSet.aSortedProteins)

    return 0

def get_args():
    parser = argparse.ArgumentParser(description='MFA')
    parser.add_argument('-f', '--file', metavar='file',
                        default='', help='CSV input file of protein purifications')
    parser.add_argument('-bp', '--bioplex', metavar='bioplex',
                        default='', help='Indicate if the input is in Bioplex format')
    parser.add_argument('-k', '--max', metavar='numclusters',
                        default='100', help='A maximum number of possible clusters')
    parser.add_argument('-psi', '--ratio', metavar='psi',
                        default='3.4', help='A ratio of log(1-fn)/log(1-fp)')
    return parser.parse_args()

def main():
    args = get_args()
    if (args.file == '' and args.bioplex == ''):
        print('Input file cannot be empty. Require a CSV file of protein purifications.')
        exit()
    
    nK = int(args.max)
    psi = float(args.ratio)

    if args.bioplex != '':
        print('Hello, ' + args.bioplex)
        inputSet = inputBioplex.CInputBioplex(args.bioplex, cpmBioplex.CountBioplexSpoke)
    else:
        print('Hello, ' + args.file)
        inputSet = inputFile.CInputSet(args.file, cpm.CountSpokeModel)
    nLogLikelihood = clustering(inputSet, nK, psi)

if __name__ == '__main__':
    main()