calc_CSP_RANK_method.py

# calc_CSP_metrics.py
# calculate Fmeasure, Precision, Recall for each apo/holo pair with CSP data
# determine binding residues using UCBShift predictions for holo form, spectra aligned to holo shifts

import sys
from os.path import basename
from util import *
from paths import *
from os.path import exists, isfile, isdir, join
from os import listdir
from tqdm import tqdm


CSPmethod = "MONTE"
# CSPmethod = "EVENAS2001"
# CSPmethod = "GRZESIEK1996"
# CSPmethod = "WILLIAMSON2013"

CSmethod = ""
while CSmethod not in ['SPARTA', 'UCBShift', 'ShiftX', 'consensus']:
    CSmethod = input("What CS prediciton method to use? [SPARTA, UCBShift, ShiftX, consensus]")

data_source_file = './csp_stats_'+CSmethod+'.csv'
if CSPmethod != "MONTE":
    data_source_file = './csp_stats_'+CSPmethod+'_'+CSmethod+'.csv'


z_value = 0

def align_shifts_to_seq(aligned_sequence, sequence, shifts):
    new_shifts = []
    seq_index = 0
    for i in range(len(aligned_sequence)):
        if seq_index < len(sequence) and aligned_sequence[i] == sequence[seq_index]:
            new_shifts.append(shifts[seq_index])
            seq_index += 1
        else:
            new_shifts.append(-1)
    return new_shifts

if __name__ == "__main__":

    data = parse_csv(data_source_file)
    apos = [str(data['apo_bmrb']) for data in data]
    bounds = [data['holo_pdb'] for data in data]
    for i, holo in enumerate(bounds):
        apo = apos[i]
        if holo in [ '1klq', '2kri', '2lgf', '2lv6', '2n7k']:
            continue
        

        print(apo)
        print(holo)
        def add_data(apo, holo):

            try:
                TP, FP, FN, TN = get_confusion(apo, holo, "MONTE", CSmethod, "", "", structure_source = "NMR_medoid")
                F, MCC, consensus = get_F_MCC_cons(TP, FP, FN, TN)
                new_values = [F, MCC, consensus]
                new_columns = ["F_NMR", "MCC_NMR", "consensus_NMR"]
                update_row(data_source_file, apo, holo, new_values, new_columns)
            except Exception as e:
                print(e)
                if e == "TypeError: cannot unpack non-iterable NoneType object":
                    print("No NMR data")
                    return
                raise

            try:
                TP, FP, FN, TN = get_confusion(apo, holo, "MONTE", CSmethod, "", "", structure_source = "AF2_top_rank")
                F, MCC, consensus = get_F_MCC_cons(TP, FP, FN, TN)
                new_values = [F, MCC, consensus]
                new_columns = ["F_AF2", "MCC_AF2", "consensus_AF2"]
                update_row(data_source_file, apo, holo, new_values, new_columns)
            except Exception as e:
                print(e)
                if e == "TypeError: cannot unpack non-iterable NoneType object":
                    print("No NMR data")
                    return
                raise
                
        try:
            add_data(apo, holo)
        except Exception as e:
            print(e)
            if e == "TypeError: cannot unpack non-iterable NoneType object":
                print("No NMR data")
                continue
            raise