Kamil

In settings->advanced I have `local:directory?type=artist` and from debug console it seems to seek for albumartist... : `Restoring "local:directory?type=artist&role=albumartist" and 109 items from database` maybe that's the case - if so,...

I think I found the reason - I was comparing hmm profiles prepared with altered neff threshold but didn't filter the a3m dataset so when hhsearch was searching the database...

Janek, I ran your command /opt/hhsuite/bin/hhblits -i query-kosinski.fst -cpu 100 -oa3m query-kosinski.a3m -n 3 -e 0.001 -maxseq 1000000 -realign_max 100000 -maxfilt 100000 -min_prefilter_hits 1000 -d /db/hh/UniRef30_2020_06 -d /home/db/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt just now...

> Hold on, or is the UniRef30_2020_06 basically updated version of uniclust30_2018_08? I believe yes - they are calling it uniclust but the files are named Uniref30... also, there's even...

Oh, and indeed, with uniclust'18 it crashed with exactly the same error as yours. :/

Another option would to merge both databases into one (which shouldn't be difficult since it is ffindex based) and modify af2 script to run it on the merged DB. Janek's...

Do you have mpirun, ffindex_apply_mpi and hhconsensus in your PATH? Check output of these commands: ``` which mpirun which ffindex_apply_mpi which hhconsensus ```

You should use full path to ffindex_apply_mpi binary. I don't know where might that be in conda... By the way - when you are using -np 1 option in mpirun...

Suppose you have files: querydb_hhm.ffdata querydb_hhm.ffindex dbToBeSearched_hhm.ffdata dbToBeSearched_hhm.ffindex You can run: `ffindex_apply querydb_hhm.ffdata querydb_hhm.ffindex -i mappings.ffindex -d mappings.ffdata -- hhsearch -i stdin -o stdout -d dbToBeSearched` and this will generate...

maybe something with 227_a3m.ff{data,index} files? You can see into the 227_a3m.ffdata file and check whether it contains anything. Another test is to run it without -v 0 option and see...