-
Notifications
You must be signed in to change notification settings - Fork 0
/
Maximize_Entropy.py
195 lines (135 loc) · 4.74 KB
/
Maximize_Entropy.py
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
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
import rdkit
from rdkit import Chem
from rdkit.Chem import AllChem
import csv
import numpy
from rdkit.Chem import DataStructs
from random import randint
from operator import itemgetter
import math
import csv
import argparse
def get_index_totals(fps):
""" Counts the number of '1'(e.g. 'on') bits at every FP index of a given mol """
indexes = [x for x in range(1024)]
idx_counts = {key: 0 for key in indexes}
ct = 0
for fp in fps:
for i,x in enumerate(fp):
if x is '1':
ct += 1
idx_counts[i] += 1
return idx_counts
def check_addition(train_fps, test_fps):
""" Checks the change in bit diversity of a given fp (test_fps) against
a set of known fps (train_fps) """
sums = 0
# Get the sum vector [1,0,12,23,....0,0,2]
train_sums = sum_fps(train_fps)
for i,x in enumerate(train_sums):
if x == 0 and test_fps[i] == 1:
sums += 1
return sums
def sum_fps(train_fps):
""" Generates a "master" fps which sums all train_fps at each index """
master_fps = numpy.zeros(1024)
for i,x in enumerate(master_fps):
index_sum = 0
for fp in train_fps:
if fp[i] == 1:
index_sum += 1
master_fps[i] = index_sum
return master_fps
def read_csv(filename):
""" Presumes input csv file is in format of ids/smiles/ic50 """
smiles = []
ic50 = []
ids = []
with open(filename) as ifile:
data = csv.reader(ifile, delimiter=',')
for line in data:
ids.append(line[0])
smiles.append(line[1])
ic50.append(float(line[2]))
return ids, smiles, ic50
def generate_fingerprints(smiles):
""" Generates ECFP4 circular fingerprints in approrpate format """
temp = []
mols = [Chem.MolFromSmiles(x) for x in smiles]
bit_info = {}
fp = [AllChem.GetMorganFingerprintAsBitVect(x, 2, nBits=1024, bitInfo=bit_info) for x in mols]
for arr in fp:
temp.append([int(x) for x in arr])
fps = numpy.array(temp, dtype=float)
return fps
def write_csv(train,test, output_name):
with open('{}-train.csv'.format(output_name), 'w') as ofile:
writer = csv.writer(ofile)
writer.writerows(train)
with open('{}-test.csv'.format(output_name), 'w') as ofile2:
writer=csv.writer(ofile2)
writer.writerows(test)
def end_condition(train, test, target_diversity, output_name):
""" Checks if end condition is met--in this case it waits until a threshold bit diversity is reached. If met, writes output files """
temp = 0
train_smiles = [x[1] for x in train]
train_fps = generate_fingerprints(train_smiles)
sums = sum_fps(train_fps)
for i,x in enumerate(sums):
if x != 0:
temp += 1
fraction = float(temp)/1024
if fraction < target_diversity:
return False
else:
write_csv(train,test, output_name)
return True
def main():
test = []
train = []
iteration_num = 1
# Argument parser
parser = argparse.ArgumentParser(description='argument parser for shannon entropy search')
parser.add_argument('--input', type=str, dest="input_file")
parser.add_argument('--output', type=str, dest="output_file")
parser.add_argument('--cutoff', type=float, dest="diversity_cutoff")
args = parser.parse_args()
input_csv_name = args.input_file
diversity_cutoff = args.diversity_cutoff
output_file = args.output_file
print ""
print "Grabbing data..."
# Grab data
ids, smiles, ic50 = read_csv(input_csv_name)
# Create training set
# For now, grabs just the last compound in the list
train.append([ids[-1],smiles[-1],ic50[-1]])
# Create the test set
num = len(smiles)
for i in range(num-1):
test.append([ids[i],smiles[i], ic50[i]])
print "Begginning search..."
while not end_condition(train, test, diversity_cutoff, output_file):
se_idxs = []
se = []
print ""
print "-"*25
print "Iteration #{}".format(iteration_num)
print "-"*25
# Generate all necessary training data
train_smiles = [x[1] for x in train]
print train_smiles
train_fps = generate_fingerprints(train_smiles)
test_smiles = [x[1] for x in test]
# Generate intial actives
test_fps = generate_fingerprints(test_smiles)
additional_bits = []
# Get all of the entropies of the test molecules
for x in test_fps:
additional_bits.append(check_addition(train_fps,x))
max_index = additional_bits.index(max(additional_bits))
train.append(test[max_index])
del test[max_index]
iteration_num += 1
if __name__ == "__main__":
main()