-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset.py
93 lines (80 loc) · 3.1 KB
/
dataset.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
import numpy as np
from torch.utils.data import Dataset
train_path = 'data/astral_train.fa'
test_path = 'data/astral_test.fa'
def fasta2dict(inf, label_mode=True):
name = None
data = {}
for line in inf:
line = line.strip()
if line.startswith('>'):
name = line.split()[0][1:]
data[name] = {}
data[name]['seq'] = ''
if label_mode:
label_split = line.split()[1].split('.')
label = label_split[0]+'.'+label_split[1]
data[name]['label'] = label
else:
data[name]['seq'] += line
return data
def load_blosum(path):
blosum = {}
with open(path, 'r') as f:
for line in f:
line = line.strip()
aa = line.split()[0]
feature = line.split()[1:]
feature = np.array(feature, dtype=float)
blosum[aa] = feature
return blosum
class SeqDataset(Dataset):
def __init__(self, opt, train=True):
self.max_length = opt.max_length
self.train = train
self.alphabet = {aa: i for i, aa in enumerate(opt.alphabet)} # XOUBZ -> *
self.blosum = load_blosum('BLOSUM62.txt')
self.data_path = train_path if self.train else test_path
self.label_map = np.loadtxt('label.txt', dtype=str)
self.class_num = self.label_map.shape[0]
with open(self.data_path, 'r') as f:
self.data = fasta2dict(f, label_mode=train)
self.length = len(self.data)
print(f"load {'train' if self.train else 'test'} dataset size: {self.length}")
def _encode_seq(self, input_seq: str):
coding = np.zeros([self.max_length]) # [PAD] = 0
coding[0] = 1 # [CLS] = 1
coding[-1] = 2 # [SEP] = 2
input_seq = input_seq[:self.max_length - 2]
for i, aa in enumerate(input_seq):
if aa in self.alphabet.keys():
coding[i + 1] = self.alphabet[aa] + 3
elif chr(ord(aa) - 32) in self.alphabet.keys():
coding[i + 1] = self.alphabet[chr(ord(aa) - 32)] + 3
else:
coding[i + 1] = self.alphabet['*'] + 3
return coding
# def _encode_label(self, input_label: int):
# output_label = [0 for _ in range(self.class_num)]
# output_label[input_label] = 1
# return np.array(output_label, dtype=int)
def __getitem__(self, item):
name = list(self.data.keys())[item]
seq = self.data[name]['seq']
if self.train:
label = self.data[name]['label']
label = int(np.where(self.label_map == label)[0])
# return name, self._encode_seq(seq), self._encode_label(label)
return name, self._encode_seq(seq), label
else:
return name, self._encode_seq(seq)
def __len__(self):
return self.length
if __name__ == '__main__':
from torch.utils.data import DataLoader
from config import opt
data = SeqDataset(opt, train=False)
dataloader = DataLoader(dataset=data, batch_size=2, shuffle=False, num_workers=1)
for n, s in dataloader:
print(s.shape)
break