ten_species.py

from typing import List
import datasets
import pandas as pd
from Bio import SeqIO

_CHUNK_LENGTHS = [16384, 32768]


def filter_fn(char: str) -> str:
    """
    Transforms any letter different from a base nucleotide into an 'N'.
    """
    if char in {'A', 'T', 'C', 'G'}:
        return char
    else:
        return 'N'


def clean_sequence(seq: str) -> str:
    """
    Process a chunk of DNA to have all letters in upper and restricted to
    A, T, C, G and N.
    """
    seq = seq.upper()
    seq = map(filter_fn, seq)
    seq = ''.join(list(seq))
    return seq


class TenSpeciesGenomesConfig(datasets.BuilderConfig):
    """BuilderConfig for The Human Reference Genome."""

    def __init__(self, *args, chunk_length: int, overlap: int = 0, **kwargs):
        """BuilderConfig for the multi species genomes.
        Args:
            chunk_length (:obj:`int`): Chunk length.
            overlap: (:obj:`int`): Overlap in base pairs for two consecutive chunks (defaults to 0).
            **kwargs: keyword arguments forwarded to super.
        """
        super().__init__(
            *args,
            name=f'{chunk_length}bp',
            **kwargs,
        )
        self.chunk_length = chunk_length
        self.overlap = overlap


class TenSpeciesGenomes(datasets.GeneratorBasedBuilder):
    """Genomes from 10 species, filtered and split into chunks of consecutive nucleotides.

    Species include:
        - Homo_sapiens
        - Mus_musculus
        - Drosophila_melanogaster
        - Danio_rerio
        - Caenorhabditis_elegans
        - Gallus_gallus
        - Gorilla_gorilla
        - Felis_catus
        - Salmo_trutta
        - Arabidopsis_thaliana
    """

    VERSION = datasets.Version("1.0.0")
    BUILDER_CONFIG_CLASS = TenSpeciesGenomesConfig
    BUILDER_CONFIGS = [TenSpeciesGenomesConfig(chunk_length=chunk_length) for chunk_length in _CHUNK_LENGTHS]
    DEFAULT_CONFIG_NAME = "32768bp"

    def _info(self):

        features = datasets.Features(
            {
                "sequence": datasets.Value("string"),
                "species_label": datasets.ClassLabel(
                    num_classes=10,
                    names=['Homo_sapiens', 'Mus_musculus', 'Drosophila_melanogaster', 'Danio_rerio',
                           'Caenorhabditis_elegans', 'Gallus_gallus', 'Gorilla_gorilla', 'Felis_catus',
                           'Salmo_trutta', 'Arabidopsis_thaliana']),
                "description": datasets.Value("string"),
                "start_pos": datasets.Value("int32"),
                "end_pos": datasets.Value("int32"),
                "fasta_url": datasets.Value("string")
            }
        )
        return datasets.DatasetInfo(
            # This defines the different columns of the dataset and their types
            features=features,
        )

    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:

        urls_filepath = dl_manager.download_and_extract('ten_species_urls.csv')
        with open(urls_filepath) as urls_file:
            all_species = [line.rstrip().split(',')[0] for line in urls_file]
        with open(urls_filepath) as urls_file:
            urls = [line.rstrip().split(',')[-1] for line in urls_file]
        all_species = tuple(all_species)
        downloaded_files = dl_manager.download_and_extract(urls)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={"all_species": all_species, "files": downloaded_files,
                            "chunk_length": self.config.chunk_length, "overlap": self.config.overlap}
            ),
        ]

    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
    def _generate_examples(self, all_species, files, chunk_length, overlap):
        key = 0
        for species, file in zip(all_species, files):
            with open(file, 'rt') as f:
                fasta_sequences = SeqIO.parse(f, 'fasta')

                for record in fasta_sequences:
                    # parse descriptions in the fasta file
                    sequence, description = str(record.seq), record.description

                    # clean chromosome sequence
                    sequence = clean_sequence(sequence)
                    seq_length = len(sequence)

                    # split into chunks
                    num_chunks = (seq_length - 2 * overlap) // chunk_length

                    if num_chunks < 1:
                        continue

                    sequence = sequence[:(chunk_length * num_chunks + 2 * overlap)]
                    seq_length = len(sequence)

                    for i in range(num_chunks):
                        # get chunk
                        start_pos = i * chunk_length
                        end_pos = min(seq_length, (i+1) * chunk_length + 2 * overlap)
                        chunk_sequence = sequence[start_pos:end_pos]

                        # yield chunk
                        yield key, {
                            'sequence': chunk_sequence,
                            'species_label': species,
                            'start_pos': start_pos,
                            'end_pos': end_pos,
                            'fasta_url': file.split('::')[-1]
                        }
                        key += 1