search

soedinglab / MMseqs2

MMseqs2: ultra fast and sensitive search and clustering suite
https://mmseqs.com
GNU General Public License v3.0
1.4k stars 195 forks source link

mmseqs taxonomy problem #80

Closed henav closed 6 years ago

henav commented 6 years ago

Hi, I am using mmseqs taxonomy, and following the taxonomy assignment workflow, using uniref100 as a target DB.

the entire workflow works great, until the "mmseqs taxonomy" command. The first search seems to be working good, however, I am getting an error in the second search, and the process dies.

MMseqs call: mmseqs taxonomy est00001DB uniref100DB uniref100DB.tsv /ncbi-taxdump/ queryLcaDB /tmp/ --threads 25

MMseqs Output

Program call: /tmp/3706626149599022609/top1 uniref100DB /tmp/3706626149599022609/top1 /tmp/3706626149599022609/aligned --extract-mode 2

MMseqs Version: 2.23394 Extract mode 2 Threads 80 Verbosity 3

Query file: /tmp/3706626149599022609/top1 Target file: uniref100DB Alignment database: /tmp/3706626149599022609/top1 Start writing file to /tmp/3706626149599022609/aligned Set sym link from /tmp/3706626149599022609/top1_h to /tmp/3706626149599022609/aligned_h Set sym link from /tmp/3706626149599022609/top1_h.index to /tmp/3706626149599022609/aligned_h.index Done. Time for merging files: 0 h 0 m 1 s Time for processing: 0 h 0 m 25s Please recreate your database or add a .dbtype file to your sequence/profile database. Error: Second search died

I am using version 2.23394 installed through conda environment, ubuntu 16.04 server Thanks for your help!

milot-mirdita commented 6 years ago

Could you please install the latest git version? I think I know whats going on, the latest commit should fix the issue.

Another thing that you can do to fix your current run is:

cp est00001DB.dbtype /tmp/3706626149599022609/top1.dbtype

And rerun the same command.

henav commented 6 years ago

Thanks for the quick reply!

I installed the latest version, now receiving another error: /tmp/814303472506594710/taxonomy.sh: 23: /tmp/814303472506594710/taxonomy.sh: abspath: not found

And the process dies immediately.

milot-mirdita commented 6 years ago

Whoops... Lastest commit fixes that. I don't have a good workaround this time for your version, I am afraid you have to recompile/reinstall MMseqs2.

henav commented 6 years ago

The version I installed is: f56650c3d80e9cb72ec9b09964d622cbbf879a60 I thought it is the latest, am I wrong?

milot-mirdita commented 6 years ago

I mean, I just had fixed your issue and commited that. There is one more commit now (3957254).

henav commented 6 years ago

Thanks for clarifying. So now I'm on version 39572549e20ed90d8e9bb4ea02fd59d6041859be . I get this error:

Query file: /tmp/2765600628860701936/top1 Target file: uniref100DB Alignment database: /tmp/2765600628860701936/top1 Start writing file to /tmp/2765600628860701936/aligned Time for merging files: 0 h 0 m 0 s Could not get realpath of /tmp/2765600628860701936/top1_h! Error: Extractalignedregion failed

milot-mirdita commented 6 years ago

I'll take a look at it tomorrow, sorry about that!

milot-mirdita commented 6 years ago

I think with the latest commit it should finally work fine. Please update to the latest version and delete the following files. Then you can rerun the previous command, it should continue and work fine now.

rm -f tmp/2765600628860701936/top1* tmp/2765600628860701936/aligned*

Thanks a lot for uncovering all these bugs!

henav commented 6 years ago

It works now, thank you for taking care of that!