This is a test version of Biostars. For the public version, visit https://www.biostars.org.
I would to like to know if there any python module to calculate a Hamming distance from multiple sequences aligment

Any of you guys have some suggestion about to calculate the HD distance and the entropy of a multiple sequence alignment?

Thanks

alignment
• 5,513 views • link
written by schlogl schlogl

I don't vouch for the contents of this repository, but its description at least matches your task: Hamming Distance Comparison of Amino Acid Sequences of 10 Organisms

• 0 views • link
written by Mensur Dlakic Mensur Dlakic

1 answer

BioPython can do all of this, but it’s pretty easy to implement yourself (and is good practise).

See for example: https://github.com/jrjhealey/bioinfo-tools/blob/master/Shannon.py

And Hamming distance is super simple: https://github.com/jrjhealey/bioinfo-tools/blob/master/StringComparisons.py#L84

• 0 views • link
written by Joe Joe

Hi Joe, I have a nice function for HD, but my doubt was about how to get all sequences checked for HD and the entropy. Once HD counts for two sequences at time. And in a MSA you have a lot of them compared. Maybe a loop checking each two sequence. I will check it out. thanks

• 0 views • link
written by schlogl schlogl

You can do all pairwise comparisons between sequences and store the numbers.

Check out the itertools module.

• 0 views • link
written by Joe Joe

maybe

itertools.imap(function, *iterables) ?

• 0 views • link
written by schlogl schlogl

I got this Joe:

def hamming_dist(s1, s2):
    assert len(s1) == len(s2)
    hd = 0
    for b1, b2 in zip(s1, s2):
        if b1 != b2:
            hd += 1
    return hd

def imap(function, *iterables):
    iterables = map(iter, iterables)
    for it in iterables:
        args = tuple(it)
        if function is None:
            yield tuple(args)
        else:
            yield function(*args) 

distances = imap(hamming_dist, *itertools.combinations(ls,2))
for dist in distances:
    print(dist)

Do you think there are some way toimprove it?

PS- Just test in a toy exaple.

• 0 views • link
written by schlogl schlogl

I think you can do this more simply if you want to use BioPython.

I forget the exact syntax now but it would be something like:

from Bio import AlignIO
import itertools
aln = AlignIO.read(...)
for r1, r2 in itertools.combinations(aln, 2):
   print("\n".join([r1.id, str(r1.seq), r2.id, str(r2.seq), str(hamming_distance(str(r1.seq), str(r2.seq)))]))

(Not the prettiest output, but you can tweak).

Your solution looks reasonable too though, so whatever works.

• 0 views • link
written by Joe Joe

I will check yours too Joe. Thanks 8)

• 0 views • link
written by schlogl schlogl

Hey Joe your solution was nicer because included the id. I will try to do that in mine. 8)

DQ676872
GCAGGAAGCACTATGGGCGCGGCGTCAATAACGCTGACGGTACAGGCCAGG---------CAATTATTGTCTGGCATAGTGCAACAGCAAAGCAATTTGCTGAGGGCTATAGAGGCTCAACAACATATGTTGAAACTCACGGTCTGGGGCATTAAACAGCTCCAGGCAAGAGTCCTGGCTCTAGAAAGATACCTAAAGGATCAACAGCTC
AB253421
GCAGGAAGCACTATGGGCGCGGCGTCAATGACGCTGACGGTACAGGCCAGA---------CAATTATTGTCTGGCATAGTGCAACAGCAAAGCAATTTGCTGAGGGCTATAGAGGCTCAACAACATCTGTTGAAACTCACGGTCTGGGGCATTAAACAGCTCCAGGCAAGAGTCCTGGCTGTGGAAAGCTACCTAAGGGATCAACAGCTC
7

Thank you for your support! Paulo

• 0 views • link
written by schlogl schlogl

Log in to answer this question.

More posts like this