check the frequency of the first N bases to try to identify barcode sequences in a non-demultiplexed fastq file

get-barcodes.py

#!/usr/bin/env python3

# usage: cat reads.fastq | ./get-barcodes.py

from Bio import SeqIO
from collections import defaultdict

BARCODE_SIZE = 14

counts = defaultdict(lambda: 0)

with open("/dev/stdin") as handle:
    for n, record in enumerate(SeqIO.parse(handle, "fastq")):
        counts[str(record.seq)[0:BARCODE_SIZE]] += 1
        if n > 100_000:
            break

for seq, count in counts.items():
    if count > 1000:
        print(seq, count)