Struct2Graph/create_examples.py at master · baranwa2/Struct2Graph · 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
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
# -*- coding: utf-8 -*-
"""
Created on Tue Jul  7 12:12:24 2020

@author: mayank
"""

import os
from os import walk

import numpy as np
from random import shuffle


filepath = "pdb_files/input1/"

for (dirpath, dirnames, filenames) in walk(filepath):
    break

os.chdir(filepath)

for word in filenames[:]:
    if not word.startswith('names'):
        filenames.remove(word)

p_list, m_list, s_list = [], [], []
for f in filenames:
    a      = np.load(f)
    st_ind = int(f.strip().split('_')[1])

    for e, p in enumerate(a):
        p_list.append(p)
        m_list.append(st_ind)
        s_list.append(e)


filepath = "../../"
os.chdir(filepath)

with open('interactions_data.txt', 'r') as f:
    protein_list = f.read().strip().split('\n')

p1_list, p2_list, int_list = [], [], []
i1_list, i2_list           = [], []
for pair in protein_list:
    p1, p2, int_val  = pair.strip().split('\t')

    if p1 in p_list and p2 in p_list:
        p1_list.append(p1)
        p2_list.append(p2)
        int_list.append(int(int_val))

        i1 = p_list.index(p1)
        i2 = p_list.index(p2)

        i1_list.append(m_list[i1]+s_list[i1]-1)
        i2_list.append(m_list[i2]+s_list[i2]-1)

### Create Examples
filepath = "pdb_files/input1/"
os.chdir(filepath)

merge_examples = []

for e, _ in enumerate(p1_list):
    merge_examples.append([i1_list[e], i2_list[e], int_list[e]])

shuffle(merge_examples)
shuffle(merge_examples)
shuffle(merge_examples)

print('Total number of examples : ' + str(len(merge_examples)))


outF = open("all_examples.txt", "w")
for line in merge_examples:
    outF.write(str(line[0]))
    outF.write("\t")
    outF.write(str(line[1]))
    outF.write("\t")
    outF.write(str(line[2]))
    outF.write("\n")
outF.close()

np.save('all_examples',merge_examples)

### Create train, test and dev examples

dataset_train_len = int(np.floor(0.8*len(merge_examples)))
dataset_rem_len   = len(merge_examples)-dataset_train_len
dataset_test_len  = int(np.floor(0.5*dataset_rem_len))

dataset_train = []
for i in range(dataset_train_len):
    dataset_train.append(merge_examples[i])

dataset_test = []
for i in range(dataset_train_len,dataset_train_len+dataset_test_len):
    dataset_test.append(merge_examples[i])

dataset_dev = []
for i in range(dataset_train_len+dataset_test_len,len(merge_examples)):
    dataset_dev.append(merge_examples[i])


np.save('train_examples',dataset_train)
np.save('one_to_one_test_examples',dataset_test)
np.save('one_to_one_dev_examples',dataset_dev)


ind_pos = []
ind_neg = []
for no, d in enumerate(dataset_train):
    if d[-1] == 1:
        ind_pos.append(no)
    else:
        ind_neg.append(no)

pos_example = len(ind_pos)


#1:2 Ratio
pos_inds = list(np.random.choice(np.array(ind_pos), int(pos_example/2), replace=False))

dataset_train_12 = []
for no, d in enumerate(dataset_train):
    if no in pos_inds or no in ind_neg:
        dataset_train_12.append(d)

shuffle(dataset_train_12)
np.save('train_examples_one_two',dataset_train_12)


#1:3 ratio
pos_inds = list(np.random.choice(np.array(ind_pos), int(pos_example/3), replace=False))

dataset_train_13 = []
for no, d in enumerate(dataset_train):
    if no in pos_inds or no in ind_neg:
        dataset_train_13.append(d)

shuffle(dataset_train_13)
np.save('train_examples_one_three',dataset_train_13)


#1:5 ratio
pos_inds = list(np.random.choice(np.array(ind_pos), int(pos_example/5), replace=False))

dataset_train_15 = []
for no, d in enumerate(dataset_train):
    if no in pos_inds or no in ind_neg:
        dataset_train_15.append(d)

shuffle(dataset_train_15)
np.save('train_examples_one_five',dataset_train_15)


#1:10 ratio
pos_inds = list(np.random.choice(np.array(ind_pos), int(pos_example/10), replace=False))

dataset_train_110 = []
for no, d in enumerate(dataset_train):
    if no in pos_inds or no in ind_neg:
        dataset_train_110.append(d)

shuffle(dataset_train_110)
np.save('train_examples_one_ten',dataset_train_110)