Closed ekg closed 4 years ago
I regenerated the index with mmseqs createdb ncbi-genomes-2020-05-03/*fna.gz seqDB
and it seems to work fine. I'm somewhat perplexed as to how it ended up in this state.
Glad it's working. That was super weird.
Regarding your guix definition. You might also want to add dependencies to bzlib and wget. (Theoretically also awk but awk seems to demanded by the POSIX standard to always be present).
You could take the homebrew recipe as a reference: https://github.com/Homebrew/homebrew-core/blob/master/Formula/mmseqs2.rb
Expected Behavior
I expected to run an alignment with
mmseqs search
and then convert the alignment into a custom tabular format withmmseqs convertalis
.Current Behavior
Running
mmseq convertails
fails withCan not open index file seqDB.index
Steps to Reproduce (for bugs)
Download refseq representative microbial genomes:
Generate the index:
Align anything, like this holliday junction resolvase:
Now, we try to extract the alignment information:
Context
Why am I missing the .index file? This is a rather large target sequence set, and so I see things are broken up into many smaller index files. But there is not one main .index.
Your Environment
This is a recent debian system (not sure which version) and I'm running a build of mmseqs that I set up in guix. https://github.com/ekg/guix-genomics/blob/master/mmseqs2.scm