-
Notifications
You must be signed in to change notification settings - Fork 1
/
pipred.py
145 lines (126 loc) · 5.03 KB
/
pipred.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
import argparse
from Bio import SeqIO
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
import sys
import random
import numpy as np
from utils import enc_seq_onehot, enc_pssm, is_fasta, get_pssm_sequence, PiPred_Model, decode,exit
import tensorflow as tf
tf.compat.v1.logging.set_verbosity(tf.compat.v1.logging.ERROR)
import tensorflow.keras.backend as K
my_loc = os.path.dirname(os.path.abspath(__file__))
parser = argparse.ArgumentParser(description='PiPred')
parser.add_argument('-i',
help='Input file with sequence in fasta format.',
required=True,
metavar='FILE')
parser.add_argument('-out_path',
help='Output directory',
default='.',
metavar='DIR')
parser.add_argument('-pssm_path',
metavar='DIR',
default='.',
help='Directory with PSSM files.')
args = parser.parse_args()
# Verify whether weights files are present
for i in range(1, 11):
if not os.path.isfile('%s/weights/final_%s.h5' % (my_loc, i)):
print("Weight files for the PiPred model are not available in weights directory.")
exit()
# INPUT VERIFICATION #
print("Veryfing input...")
# Check if input file exists
if not os.path.isfile(args.i):
print('ERROR: Input file does not exist!')
exit()
# Check if input is valid fasta file
if not is_fasta(args.i):
print("ERROR: Malformed fasta file. Please check input!")
exit()
if not os.path.isdir(args.out_path):
print("ERROR: Output directory does not exist!")
exit()
# Parse fasta file
input_data = list(SeqIO.parse(args.i, "fasta"))
sequences = [str(data.seq) for data in input_data]
entries = [''.join(e for e in str(data.id) if (e.isalnum() or e == '_')) for data in input_data]
if not len(entries) == len(set(entries)):
print("ERROR: Sequence identifiers in the fasta file are not unique!")
exit()
# Check sequence length and presence of non standard residues
aa1 = "ACDEFGHIKLMNPQRSTVWYX"
for entry, seq in zip(entries, sequences):
if not (len(seq) >= 25 and len(seq) <= 700):
print('ERROR: Not accepted sequence length (ID %s - %s). Only sequences between 30 and 700 residues are accepted!' % (
entry, len(seq)))
exit()
for aa in seq:
if aa not in aa1:
print("ERROR: Sequence (ID %s) contains non-standard residue (%s)." % (entry, aa))
exit()
# PSSM SPECIFIC INPUT VERIFICATION #
pssm_files = []
# Check if directory exists
if not os.path.isdir(args.out_path):
print("ERROR: Directory with PSSM files does not exist!")
exit()
for entry, seq in zip(entries, sequences):
pssm_fn = '%s/%s.pssm' % (args.pssm_path, entry)
if not os.path.isfile(pssm_fn):
print("ERROR: PSSM file for entry %s does not exist!" % entry)
exit()
if not get_pssm_sequence(pssm_fn) == seq:
print("ERROR: Sequence in PSSM file does not match fasta sequence for entry %s!" % entry)
exit()
try:
parsed_pssm = np.genfromtxt(pssm_fn, skip_header=3, skip_footer=5, usecols=(i for i in range(2, 22)))
if parsed_pssm.shape[0] == len(seq) - 2:
parsed_pssm = np.genfromtxt(pssm_fn, skip_header=3, skip_footer=3, usecols=(i for i in range(2, 22)))
elif parsed_pssm.shape[0] != len(seq):
raise ValueError
except ValueError:
print("ERROR: Malformed PSSM file for entry %s!" % entry)
exit()
pssm_files.append(pssm_fn)
print("Encoding sequences...")
# Encode sequence into vector format
enc_sequences = []
for seq, pssm_fn in zip(sequences, pssm_files):
pad_left = random.randint(0, 700 - len(seq))
enc_sequences.append(np.concatenate((enc_seq_onehot(seq, pad_length=700),
enc_pssm(pssm_fn, pad_length=700)), axis=1))
# Create model
model = PiPred_Model()
enc_sequences = np.asarray(enc_sequences)
ensemble_results = {}
print("Predicting...")
for i in range(1, 11):
model.load_weights('%s/weights/final_%s.h5' % (my_loc, i))
predictions = model.predict(enc_sequences)
decoded_predictions = [decode(pred, encoded_seq) for pred, encoded_seq in
zip(predictions, enc_sequences)]
for decoded_prediction, entry in zip(decoded_predictions, entries):
if i == 1:
ensemble_results[entry] = []
k = 0
for aa_col in decoded_prediction.T:
if i == 1:
ensemble_results[entry].append(aa_col)
else:
ensemble_results[entry][k] = np.vstack((ensemble_results[entry][k], aa_col))
k += 1
K.clear_session()
# Dump the results
for entry, seq in zip(entries, sequences):
f = open('%s/%s.out' % (args.out_path, entry), 'w')
f.write('aa prH prI prE prC\n')
final_results = np.asarray([np.average(ss_col, axis=0) for ss_col in ensemble_results[entry]])
for aa, probs in zip(seq, final_results.T):
f.write("%s " % aa)
for prob in probs:
f.write(" %.3f" % prob)
f.write("\n")
f.close()
print("Done!")