Non-indexed FASTA: access sequence by genomic region

[apraga]

Hi,

Indexed FASTA have extract to quickly get a sequence. I could not find a way to do the same without an index. Could such a method be added ?

Expected Behavior

For example, with sequence("chr1", 10, 1000, reader)

Possible Solution / Implementation

My attempt was

function sequence(chrom, x, y, reader)
    r= first(reader)
    seq = nothing
    while iterate(reader) != nothing
        if identifier(r) == chrom
            seq = FASTX.sequence(r)[x:y]
            break
        end
    end
    return seq
end

Your Environment

• Package Version used: 2.1.2
• Julia Version used: 1.9.2
• Operating System and version (desktop or mobile): Gentoo

Thanks,

[jakobnissen]

Dear @apraga

Thanks for making the issue and showing interest in FASTX. I don't think this is a good idea for a few reasons:

1. It's easy to implement yourself:

   julia> sequence(first(i for i in reader if identifier(i) == "S6C294"))[100:200]
   "CGACTTCGGGCCGATGTTGTTTCAATGCCGTCTGAATCGCTTCCATCCGGATTTCGGCAAGCAGGTTCACGCCTTCTGTGGGCAGTTCCAAACGTTGCGCG"

2. The code above generalizes better - you can pick the first sequence that fulfills any arbitrary function, not just from its identifier.
3. In the above code, it's clear that the operation is O(N), whereas extract is O(1), and it's also clear it's not random access which extract is.