tpierrot commited on
Commit
b20a52c
1 Parent(s): 6a2682d

Update human_reference_genome.py

Browse files
Files changed (1) hide show
  1. human_reference_genome.py +30 -11
human_reference_genome.py CHANGED
@@ -16,7 +16,6 @@
16
 
17
  from typing import List
18
  import datasets
19
- import gzip
20
  from Bio import SeqIO
21
  import regex as re
22
 
@@ -49,7 +48,7 @@ _URLS = {
49
  f"fasta": "https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/001/405/GCF_000001405.40_GRCh38.p14/GCF_000001405.40_GRCh38.p14_genomic.fna.gz"
50
  }
51
 
52
- _CHUNK_LENGTH = 6100
53
  _OVERLAP = 100
54
  _THRESHOLD_FILTER_N = 0.05
55
 
@@ -93,12 +92,33 @@ def continue_loop(split: str, chromosome: str) -> bool:
93
  return False
94
 
95
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
96
  class HumanReferenceGenome(datasets.GeneratorBasedBuilder):
97
  """Human reference genome, filtered and split into chunks of consecutive
98
  nucleotides. The test set corresponds to chromosome 22, the validation set to
99
  chromosome 21 and all other chromosomes are used for training."""
100
 
101
  VERSION = datasets.Version("1.1.0")
 
 
 
102
 
103
  def _info(self):
104
 
@@ -128,13 +148,13 @@ class HumanReferenceGenome(datasets.GeneratorBasedBuilder):
128
  downloaded_files = dl_manager.download_and_extract(urls_to_download)
129
 
130
  return [
131
- datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files['fasta'], "split": "train"}),
132
- datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files['fasta'], "split": "validation"}),
133
- datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files['fasta'], "split": "test"}),
134
  ]
135
 
136
  # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
137
- def _generate_examples(self, filepath, split):
138
  with open(filepath, 'rt') as f:
139
  fasta_sequences = SeqIO.parse(f, 'fasta')
140
  # regex to filter lines of interest in the FASTA
@@ -160,14 +180,14 @@ class HumanReferenceGenome(datasets.GeneratorBasedBuilder):
160
  seq_length = len(sequence)
161
 
162
  # split into chunks
163
- num_chunks = (seq_length - 2 * _OVERLAP) // _CHUNK_LENGTH
164
- sequence = sequence[:(_CHUNK_LENGTH * num_chunks + 2 * _OVERLAP)]
165
  seq_length = len(sequence)
166
 
167
  for i in range(num_chunks):
168
  # get chunk
169
- start_pos = i * _CHUNK_LENGTH
170
- end_pos = min(seq_length, (i+1) * _CHUNK_LENGTH + 2 * _OVERLAP)
171
  chunk_sequence = sequence[start_pos:end_pos]
172
 
173
  # compute ratio of Ns
@@ -182,4 +202,3 @@ class HumanReferenceGenome(datasets.GeneratorBasedBuilder):
182
  'end_pos': end_pos
183
  }
184
  key += 1
185
-
 
16
 
17
  from typing import List
18
  import datasets
 
19
  from Bio import SeqIO
20
  import regex as re
21
 
 
48
  f"fasta": "https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/001/405/GCF_000001405.40_GRCh38.p14/GCF_000001405.40_GRCh38.p14_genomic.fna.gz"
49
  }
50
 
51
+ _CHUNK_LENGTHS = [6000, 12000]
52
  _OVERLAP = 100
53
  _THRESHOLD_FILTER_N = 0.05
54
 
 
92
  return False
93
 
94
 
95
+ class HumanReferenceGenomeConfig(datasets.BuilderConfig):
96
+ """BuilderConfig for The Human Reference Genome."""
97
+
98
+ def __init__(self, *args, chunk_length: int, **kwargs):
99
+ """BuilderConfig for The Pile.
100
+ Args:
101
+ chunk_length (:obj:`int`): Chunk length.
102
+ **kwargs: keyword arguments forwarded to super.
103
+ """
104
+ num_kbp = int(chunk_length/1000)
105
+ super().__init__(
106
+ *args,
107
+ name="+".join(f'{num_kbp}kbp'),
108
+ **kwargs,
109
+ )
110
+ self.chunk_length = chunk_length
111
+
112
+
113
  class HumanReferenceGenome(datasets.GeneratorBasedBuilder):
114
  """Human reference genome, filtered and split into chunks of consecutive
115
  nucleotides. The test set corresponds to chromosome 22, the validation set to
116
  chromosome 21 and all other chromosomes are used for training."""
117
 
118
  VERSION = datasets.Version("1.1.0")
119
+ BUILDER_CONFIG_CLASS = HumanReferenceGenomeConfig
120
+ BUILDER_CONFIGS = [HumanReferenceGenomeConfig(chunk_length=chunk_length) for chunk_length in _CHUNK_LENGTHS]
121
+ DEFAULT_CONFIG_NAME = "6kbp"
122
 
123
  def _info(self):
124
 
 
148
  downloaded_files = dl_manager.download_and_extract(urls_to_download)
149
 
150
  return [
151
+ datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files['fasta'], "split": "train", "chunk_length": self.config.chunk_length}),
152
+ datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files['fasta'], "split": "validation", "chunk_length": self.config.chunk_length}),
153
+ datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files['fasta'], "split": "test", "chunk_length": self.config.chunk_length}),
154
  ]
155
 
156
  # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
157
+ def _generate_examples(self, filepath, split, chunk_length):
158
  with open(filepath, 'rt') as f:
159
  fasta_sequences = SeqIO.parse(f, 'fasta')
160
  # regex to filter lines of interest in the FASTA
 
180
  seq_length = len(sequence)
181
 
182
  # split into chunks
183
+ num_chunks = (seq_length - 2 * _OVERLAP) // chunk_length
184
+ sequence = sequence[:(chunk_length * num_chunks + 2 * _OVERLAP)]
185
  seq_length = len(sequence)
186
 
187
  for i in range(num_chunks):
188
  # get chunk
189
+ start_pos = i * chunk_length
190
+ end_pos = min(seq_length, (i+1) * chunk_length + 2 * _OVERLAP)
191
  chunk_sequence = sequence[start_pos:end_pos]
192
 
193
  # compute ratio of Ns
 
202
  'end_pos': end_pos
203
  }
204
  key += 1