The fastest protein sequence aligner available

Question

[Bioaln]

Hello all, I've been recently working with sequence alignment of proteins. I was wondering (I am still new to this field), what is the current state-of-the-art in protein sequence alignment?

Which are some of the fastest protein-protein alignment algorithms, faster than blastp?

Thank!

[lieven.sterck]

quick comment: do NOT consider blastP as a protein aligner (or any kind of aligner for that manner).

Blast is a search tool (aka sequence-google), not an alignment tool!

[Bioaln]

Why not? It works by protein sequence alignment, does it not?

[Joe]

It is a local aligner, but that is very different from doing, for example, a multiple sequence alignment (even if the MSA does local alignment).

[lieven.sterck]

apart from the reason pointed out by jrj.healey , blast is designed (and optimized) to look for similarities between an input protein and a database of proteins based on (and this is where the confusion comes from) locally aligned similarities.

Where alignment is a means for blast it is not the goal, while it is the goal of an alignment tool.

BLAST : Basic Local Alignment SEARCH TOOL

If you would do a blast2seq run (comparing 2 prots to each other) you might get away with it, but still I would not cataloge blast as an alignment tool.

[cpad0112]

Muscle, t-coffee.. I am not sure if they are fast. But they are supposed to be better. Use multiple aligners at same time: m-coffee from t-coffee development group.

Answer 1

[lieven.sterck]

From the top of my head (and assuming you are looking for global aligners): muscle, EMBOSS-emma, MAFFT, clustal-omega, MEGA (?) ... likely forgetting several others as well

[Bioaln]

Thanks, MAFFT seems promising!

[Joe]

+1 for pretty much all the above suggestions. Muscle in particular is crazy fast.

Clustal Omega gives some of the most accurate protein alignments to my knowledge, even if its a little slower, so it depends what you're trying to do/show.

Answer 2

[lakhujanivijay]

Also adding diamond to the list

• link

• paper

• manual

[lieven.sterck]

Do not fully agree on this: same comment/issue as with blast : it's a search tool (and even doing a worse job than blast on the alignment part)

But I do like Diamond otherwise!

[Alexander]

Notebook example: https://www.kaggle.com/code/geraseva/diamond