data_loader/data_loader.py at master · deepphage/data_loader · 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
import torch
from os import listdir
import re
from itertools import product
import numpy as np
from os.path import isfile, join
from torch.utils.data import Dataset, DataLoader, ConcatDataset
from data_loader.DnaLoad import DnaLoad

from dna2vec.dna2vec.multi_k_model import MultiKModel

class DnaDataSet(Dataset):

	def __init__(self, reads, read_labels):
		self.reads = reads
		self.read_labels = read_labels

	def __len__(self):
		return len(self.read_labels)


	def __getitem__(self, index):
		X = self.reads[index]
		y = self.read_labels[index]
		return X, y

class PhageFileLoader():
	def __init__(self, file):
		self.file = file
		self.dna_load = DnaLoad(file)

	def get_kmers_for_read(self, k, stride, read_n, read_length=100, embedding=None, embed_size=None):
		kmers = self.dna_load.get_kmer(k, stride, read_n, embedding, embed_size)
		labels = self.dna_load.get_labels(k, stride, read_n)
		# split into reads of length read_length
		reads = kmers[:read_length * (kmers.shape[0] // read_length)].reshape(-1, read_length)
		read_labels = labels[:read_length * (labels.shape[0] // read_length)].reshape(-1, read_length)
		return DnaDataSet(reads, read_labels)

	def get_kmers_for_read_w_id(self, k, stride, read_n, read_length=100, embedding=None, embed_size=None):
		kmers = self.dna_load.get_kmer(k, stride, read_n, embedding, embed_size)
		labels = self.dna_load.get_labels(k, stride, read_n)
		k_id = self.dna_load.get_k_id(read_n)
		# split into reads of length read_length
		reads = kmers[:read_length * (kmers.shape[0] // read_length)].reshape(-1, read_length)
		read_labels = labels[:read_length * (labels.shape[0] // read_length)].reshape(-1, read_length)
		return (DnaDataSet(reads, read_labels), k_id)

	def get_n_loaders(self, n='all', read_length=100, batch_size=32, k=3, stride=1, embedding=None, embed_size=None):
		loaders = []
		if(n=="all"):
			n = self.dna_load.get_number_genomes()
		datasets = np.empty(n, dtype=object)
		for i in range(n):
			dataset = self.get_kmers_for_read(k, stride, i, read_length, embedding=embedding, embed_size=embed_size)
			datasets[i] = dataset
		d = ConcatDataset(datasets)
		return d

	def get_n_loaders_w_id(self, n='all', read_length=100, batch_size=32, k=3, stride=1, embedding=None, embed_size=None):
		loaders = []
		if(n=="all"):
			n = self.dna_load.get_number_genomes()
		k_ids = []
		datasets = np.empty(n, dtype=object)
		for i in range(n):
			(dataset, k_id) = self.get_kmers_for_read_w_id(k, stride, i, read_length, embedding=embedding, embed_size=embed_size)
			_k_ids = [k_id]*len(dataset)
			k_ids.extend(_k_ids)
			datasets[i] = dataset
		d = ConcatDataset(datasets)
		# print("Lenghts of 1 file")
		# print(len(d))
		# print(len(k_ids))
		return (d, k_ids)


class PhageLoader():
	def __init__(self, datafolder):
		self.datafolder = datafolder
		self.filenames = self.get_file_names()
		self.phageLoaders = np.empty(len(self.filenames), dtype=object)

	def get_file_names(self):
		files = listdir(self.datafolder)
		regex = re.compile('Input*')
		selected_files = list(filter(regex.search, files))
		return selected_files

	def get_data_loader(self, n_files='all' ,n='all', read_length=100, batch_size=32, k=3, stride=1, embedding=None, embed_size=None, drop_last=False):
		if(n_files=='all'):
			n_files = len(self.filenames)
		datasets = np.empty(n_files, dtype=object)
		for i in range(n_files):
			dl = PhageFileLoader(self.datafolder+self.filenames[i])
			dataset = dl.get_n_loaders(n=n,read_length=read_length, batch_size=batch_size,k=k,stride=stride,embedding=embedding,embed_size=embed_size)
			datasets[i] = dataset
		d = ConcatDataset(datasets)
		return DataLoader(dataset=d, batch_size=batch_size, drop_last=drop_last)


	def get_data_set(self, n_files='all' ,n='all', read_length=100, batch_size=32, k=3, stride=1, embedding=None, embed_size=None, drop_last=False):
		if(n_files=='all'):
			n_files = len(self.filenames)
		datasets = np.empty(n_files,  dtype=object)
		for i in range(n_files):
			dl = PhageFileLoader(self.datafolder+self.filenames[i])
			dataset = dl.get_n_loaders(n=n,read_length=read_length, batch_size=batch_size,k=k,stride=stride,embedding=embedding,embed_size=embed_size)
			datasets[i] = dataset
		d = ConcatDataset(datasets)
		return d

	def get_data_set_ids(self, n_files='all' ,n='all', read_length=100, batch_size=32, k=3, stride=1, embedding=None, embed_size=None, drop_last=False):
		if(n_files=='all'):
			n_files = len(self.filenames)
		datasets = np.empty(n_files,  dtype=object)
		k_ids = []
		for i in range(n_files):
			dl = PhageFileLoader(self.datafolder+self.filenames[i])
			(dataset, k_id) = dl.get_n_loaders_w_id(n=n,read_length=read_length, batch_size=batch_size,k=k,stride=stride,embedding=embedding,embed_size=embed_size)
			k_ids.extend(k_id)
			datasets[i] = dataset
		d = ConcatDataset(datasets)
		# print(len(d))
		# print(len(k_ids))
		return k_ids

	def get_dict(self, window_size, embedding='dict'):
		letters = 'AGCT'
		vocab = [''.join(i) for i in product(letters, repeat = window_size)]
		kmers_dict = {}
		for i in range(len(vocab)):
			kmers_dict[vocab[i]] = i
		if(embedding=='dna2vec'):
			print("loading DNA 2 vec model")
			filepath = 'dna2vec/pretrained/dna2vec-20161219-0153-k3to8-100d-10c-29320Mbp-sliding-Xat.w2v'
			mk_model = MultiKModel(filepath)
			for kmer in kmers_dict.keys():
				kmers_dict[kmer] = mk_model.vector(kmer)
		return kmers_dict