MpraScreen/setup_project_data.py at main · fl16180/MpraScreen · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
import argparse
import os
import pandas as pd

from constants import *
from utils.bed_utils import *
from utils.data_utils import *
from datasets import load_data_set

SPLIT_CHOICES = [None, 'train-test', 'all']


def safe_make_dir(directory):
    if not os.path.exists(directory):
        os.makedirs(directory)


def setup(args):

    project_dir = PROCESSED_DIR / args.project

    safe_make_dir(project_dir / 'models')
    safe_make_dir(project_dir / 'neighbors')
    safe_make_dir(project_dir / 'output')

    # load or create variants bedfile
    try:
        if args.bed:
            raise Exception('Overwriting bedfile')
        bedfile = load_bed_file(args.project)
        print(f'Loaded bedfile from {args.project}')
    except Exception as e:
        if 'unif_background' in args.project:
            bedfile = get_random_bed(n_samples=100000)
        elif '1kg_background' in args.project:
            bedfile = sample_1kg_background(n_samples=100000)
        else:
            bedfile = get_bed_from_mpra(args.project)
        save_bed_file(bedfile, args.project)
        print(f'Generated new bedfile in {args.project}')

    if args.roadmap:
        print('Extracting Roadmap: ')
        fname = project_dir / 'roadmap_extract.tsv'
        extract_roadmap(bedfile.copy(), fname, args.project)

    if args.regbase:
        print('Extracting regBase: ')
        fname = project_dir / 'regBase_extract.tsv'
        extract_regbase(bedfile.copy(), fname)

    if args.eigen:
        print('Extracting Eigen: ')
        fname = project_dir / 'eigen_extract.tsv'
        extract_eigen(bedfile.copy(), fname)

    if args.genonet:
        print('Extracting GenoNet: ')
        fname = project_dir / f'GenoNet_extract.tsv'
        extract_genonet(bedfile.copy(), fname, 'mean')

    # further process data and split into train/test sets
    if args.split == 'train-test':
        bed_train, bed_test = split_train_test(bedfile,
                                               test_frac=0.25,
                                               seed=args.seed)

        process_datasets(args, bed_train, split='train')
        process_datasets(args, bed_test, split='test')

    elif args.split == 'all':
        process_datasets(args, bedfile, split='all')


def process_datasets(args, bedfile, split='all', merge='inner'):
    print(f'Processing {split} set')
    project_dir = PROCESSED_DIR / args.project

    if 'background' in args.project or (args.project == 'genomeScreen' or 'gs_job' in args.project):
        mpra = bedfile.copy()
        mpra['Label'] = 0
    else:
        mpra = load_mpra_data(args.project)

    if args.project == 'mpra_nova':
        mpra['Label'] = mpra['Hit']

    y_split = pd.merge(bedfile, mpra, on=['chr', 'pos'], how=merge).loc[:, ['chr', 'pos', 'Label']]
    y_split.to_csv(project_dir / f'{split}_label.csv', sep=',', index=False)

    if os.path.exists(project_dir / 'roadmap_extract.tsv'):
        print('\tRoadmap')
        roadmap = pd.read_csv(project_dir / 'roadmap_extract.tsv', sep='\t')
        try:
            roadmap['chr'] = roadmap['chr'].map(lambda x: x[3:])
        except TypeError:
            pass
        roadmap[['chr', 'pos']] = roadmap[['chr', 'pos']].astype(int)
        r_split = pd.merge(bedfile[['chr', 'pos']], roadmap, on=['chr', 'pos'], how=merge)
        r_split.to_csv(project_dir / f'{split}_roadmap.csv', sep=',', index=False)

    if os.path.exists(project_dir / 'regBase_extract.tsv'):
        print('\tregBase')
        regbase = clean_regbase_data(project_dir / 'regBase_extract.tsv')
        r_split = pd.merge(bedfile[['chr', 'pos']], regbase, how=merge)
        r_split.to_csv(project_dir / f'{split}_regbase.csv', index=False)

    if os.path.exists(project_dir / 'eigen_extract.tsv'):
        print('\tEigen')
        eigen = clean_eigen_data(project_dir / 'eigen_extract.tsv')
        e_split = pd.merge(bedfile[['chr', 'pos']], eigen, how=merge)
        e_split.to_csv(project_dir / f'{split}_eigen.csv', index=False)

    # compile data splits into single csv
    _ = load_data_set(args.project, split=split, make_new=True)


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--project', '-p', required=True)
    parser.add_argument('--bed', '-b', default=False, action='store_true',
                        help='(re-)extract variant bed from target data')
    parser.add_argument('--roadmap', '-r', default=False, action='store_true',
                        help='extract Roadmap data')
    parser.add_argument('--regbase', '-rb', default=False, action='store_true',
                        help='extract regBase data')
    parser.add_argument('--eigen', '-e', default=False, action='store_true',
                        help='extract Eigen data')
    parser.add_argument('--genonet', '-g', default=False, action='store_true',
                        help='extract GenoNet data')
    parser.add_argument('--split', '-s', default=None, choices=SPLIT_CHOICES,
                        help='split data into train/test sets or all')
    parser.add_argument('--seed', default=9999, help='train/test random seed')
    args = parser.parse_args()

    setup(args)