scripts/01_nullomer_identification.py

import os
import gzip
import time
import requests
from itertools import product
from Bio import SeqIO
from Bio.Seq import Seq
import numpy as np

RESULTS_DIR = "results"
DATA_DIR = "data"
K = 11
GENOME_URL = "https://ftp.ensembl.org/pub/release-110/fasta/saccharomyces_cerevisiae/dna/Saccharomyces_cerevisiae.R64-1-1.dna.toplevel.fa.gz"

os.makedirs(RESULTS_DIR, exist_ok=True)
os.makedirs(DATA_DIR, exist_ok=True)


def download_genome(genome_file):
    gz_file = genome_file + ".gz"
    if not os.path.exists(genome_file):
        response = requests.get(GENOME_URL, timeout=300, stream=True)
        if response.status_code != 200:
            raise RuntimeError(f"Genome download failed: HTTP {response.status_code}")
        with open(gz_file, "wb") as f:
            for chunk in response.iter_content(chunk_size=8192):
                f.write(chunk)
        with gzip.open(gz_file, "rt") as f_in, open(genome_file, "w") as f_out:
            f_out.write(f_in.read())
        os.remove(gz_file)


def load_genome(genome_file):
    records = list(SeqIO.parse(genome_file, "fasta"))
    whole_genome = "".join(str(r.seq).upper() for r in records)
    return whole_genome, records


def find_present_kmers(sequence, k):
    present = set()
    for i in range(len(sequence) - k + 1):
        kmer = sequence[i:i + k]
        if all(b in "ACGT" for b in kmer):
            present.add(kmer)
            present.add(str(Seq(kmer).reverse_complement()))
    return present


def identify_nullomers(present_kmers, k):
    bases = ["A", "C", "G", "T"]
    theoretical = set("".join(p) for p in product(bases, repeat=k))
    return theoretical - present_kmers, theoretical


def compute_gc(seq):
    return (seq.count("G") + seq.count("C")) / len(seq)


def main():
    genome_file = os.path.join(DATA_DIR, "yeast_genome.fsa")
    download_genome(genome_file)
    whole_genome, _ = load_genome(genome_file)

    t0 = time.time()
    present_kmers = find_present_kmers(whole_genome, K)
    nullomers, theoretical = identify_nullomers(present_kmers, K)
    pct = 100 * len(nullomers) / len(theoretical)

    sample = list(nullomers)[:10000]
    mean_gc = np.mean([compute_gc(s) for s in sample])

    out_path = os.path.join(RESULTS_DIR, f"nullomers_k{K}.txt")
    with open(out_path, "w") as f:
        for nm in sorted(nullomers):
            f.write(nm + "\n")

    print(f"Nullomers: {len(nullomers):,} ({pct:.2f}%)  GC: {mean_gc*100:.1f}%  "
          f"time: {time.time()-t0:.1f}s")


if __name__ == "__main__":
    main()