soedinglab / hh-suite

Remote protein homology detection suite.
https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-019-3019-7
GNU General Public License v3.0
529 stars 133 forks source link

Pairwise MSA with hhalign #257

Closed LeonardoMorenoG closed 3 years ago

LeonardoMorenoG commented 3 years ago

Context

I am trying to use hhalign to do a pairwise alignment of two MSA using the following command: hhalign -M 50 -t ma1.fasta -M 50 -i clustersAlignments/cluster_114.muscle.fasta -o ma2.hhr -Ofas ma2.fasta -hide_cons -glob

Expected Behavior

I expected that my output fasta included 5 sequences (2 from the file ma1.faa and 3 from the file cluster_114.muscle.fasta)

Current Behavior

Instead, the output fasta file only includes 2 sequences (1 from each file)

Question

Am I wrong when I expect as output a MSA with 5 sequences?

milot-mirdita commented 3 years ago

You have to call hhalign with -oa3m and use reformat.pl to get a non-pairwise FASTA file. If you compile from source, hhalign recently learned the -all flag to also get the unfiltered output with all sequences, otherwise you will only get a filtered MSA from hhalign.

LeonardoMorenoG commented 3 years ago

Thanks for the information. It worked perfectly. I'll close the issue

alephreish commented 2 years ago

@milot-mirdita The new -all flag is a very important addition. (BTW I tried to mark all sequences with @ and use the -mark flag, but it is simply ignored.) Is there a time frame for the next release? The version available via conda doesn't have a way to preserve all the sequence.