HMMER: hmmscan ---> hmmsearch

Question

[Metaxo]

I'm using HMMER to find domains to a battery of viral contigs (in fasta file) against Pfam database. I'm not using MSAs because most of the contigs are not related so cannot be aligned. I used hmmscan first but the running speed is just too slow. I wanted to use hmmsearch but the query file needs to be in hmm profile format in order to fit in the hmmsearch program. If I were to use hmmbuild to convert this fasta file to hmm, it would require it to be MSA. So I'm kinda stuck in here.

So.. here's what it is : Is there anyway i can end up using hmmsearch instead of hmmscan for a non-MSA fasta file?

[Joe]

You could switch to HHSuite instead.

[Metaxo]

Will that be faster than the hmmscan?

[Joe]

I don’t have speed comparisons to hand but HHsuite is fast.

It will accept a list of fastas as input though, so it solves your input issue. If you’re using long sequences I think it will be reasonably slow no matter what you do, it’s more common to compare gene size sequences.

[Metaxo]

Thanks! I will try that right now

Answer 1

[h.mon]

I know the question is a bit old, but I didn't see it before. You don't need to convert the fasta files to hmms, you can certainly use hmmsearch with the viral contigs (but remember, HMMER does protein-protein comparisons or DNA-DNA comparisons, but not DNA-protein comparisons). The format of the command-line is the same:

# hmmscan
hmmscan Pfam.hmm myorfs.fa

# hmmsearch
 hmmsearch Pfam.hmm myorfs.fa

The post hmmscan vs. hmmsearch speed: the numerology explains in detail when to use one or the other, and it states:

hmmscan and hmmsearch are doing exactly the same compute, at heart: comparing one profile to one sequence at a time. Their bit score results are identical. You can save hmmsearch tabular output files and use ’em just the same way you were going to use the hmmscan files. (Um, watch out for E-values: remember that E-values depend on the size of the database you search.)