Consensus Sequence

[j23414]

Any thoughts on smof being able to output a consensus sequence when given an alignment file?

input fasta

>seq_1
AAATATAT
>seq_2
AAAATTAT
>seq_3
AAATATAA

output fasta

>Consensus
AAATATAT

[arendsee]

@j23414 Thanks for the suggestion. It's a fine idea and I think falls within the intended scope of smof. I'll put it on the TODO list.

[arendsee]

I've partially implemented a consensus function over on the dev branch. You can check out the comment in the commit message.

The current output is a table of character counts across the columns in the alignment. But if we want to go from this table to a single consensus string, we will need to decide how to resolve ties.

[j23414]

For ties, would it make sense to output both options (both consensus sequences)?

Although I could see that becoming many consensus sequences (ties in more than 1 position)...

[arendsee]

I can't do that in FASTA format.

Perhaps if there is a near tie, I could use a wildcard, like *. For cutoff, I could use Shannon entropy. This is the standard measure conservation used in LOGO plots and the like.

[j23414]

As a standard input for LOGO plots would be acceptable.

By multiple consensuses I'm talking about:

>consensus_1
AATATAT
>consensus_2
AAAATAT

I'm looking at influenza virus sequence which group as H1.alpha, H1.beta, H1.gamma, etc...
I need the consensus for each group (alpha, beta, gamma... ) and am aligning their consensus(es) so I can determine their major nucleotide (or amino acid) differences between groups.

This might explain why a vaccine works for alpha (targets certain amino acids), but not for gamma, etc...

No worries either way, I can work with the table of character counts. : )

[arendsee]

Hmm, I'd worry about a combinatorial explosion.

[arendsee]

@j23414 You can check out the lastest version (2.13.0) of smof. Now the consensus command prints the consensus (resolving ties alphabetically) by default. Printing a table is optional.

Is good?

[j23414]

Good! : ) Consensus and table look good.

smof consensus fasta_aln.fna > consensus.fna
smof consensus -t fasta_aln.fna > consensus.tsv