This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Erorr running HHblits

Hi all,

I'm trying to run HHblits but running into an error. I have several multiple sequence alignments that I have trimmed and concatenated and then converted the resulting .phy file to a .fasta (so HHblits can accept it).

Downloaded the database using:

wget http://www.gwdg.de/~compbiol/uniclust/2018_08/uniclust30_2018_08_hhsuite.tar.gz
tar xzvf uniclust30_2018_08_hhsuite.tar.gz

When I run HHblits though, I keep getting this error message:

$ hhblits -i sample.fasta -o hhblits_results.hhr -d ./databases/uniclust30_2018_08

- 13:51:18.331 ERROR: In /dev/shm/ebuser/avx2/HHsuite/3.0-beta.3/iccifort-2018.3/hh-suite-3.0-beta.3/src/hhdatabase.cpp:35: FFindexDatabase:
- 13:51:18.331 ERROR:   could not open file './databases/uniclust30_2018_08_cs219.ffdata'

my ./databases/uniclust30_2018_08 contains the following files:

uniclust30_2018_08_a3m_db uniclust30_2018_08.cs219 uniclust30_2018_08_hhm_db uniclust30_2018_08_md5sum uniclust30_2018_08_a3m_db.index uniclust30_2018_08_cs219.ffdata uniclust30_2018_08_hhm_db.index uniclust30_2018_08_a3m.ffdata uniclust30_2018_08_cs219.ffindex uniclust30_2018_08_hhm.ffdata uniclust30_2018_08_a3m.ffindex uniclust30_2018_08.cs219.sizes uniclust30_2018_08_hhm.ffindex

Thank you

protein homology hmm

I have a similar question as the author of this post at - https://github.com/soedinglab/hh-suite/issues/167

Which link did you download your hhblits database from, and how are they different for the hhblits runs / results? uniclust30_2018_08_hhsuite.tar.gz = 25G, @ http://gwdu111.gwdg.de/~compbiol/uniclust/2018_08/uniclust30_2018_08_hhsuite.tar.gz OR uniclust30_2018_08.tar.gz = 6.6G, @ http://gwdu111.gwdg.de/~compbiol/uniclust/2018_08/uniclust30_2018_08.tar.gz

Since you are running hhblits successfully, I thought you might know :) Could you help please? Thanks!

2 answers

In the older versions at least, you had to provide the base name of the file itself, not just the directory.

With hhsearch its sufficient to use the /path/to/databasename/databasename.ffdata. With hhblits I believe it needs more than a single file, so you provide the basename of the files.

In your case it may look something like ./databases/uniclust30_2018_08/uniclust30_2018_08.

This is also assuming you’re in the correct directory, since you’re using relative filepaths.

That fixed it! I just needed to provide the basename of the files. Thanks very much!

No problem. Learned from bitter experience :P

try to give the full path of the query sequence as well as database

I have a similar question as the author of this post at - https://github.com/soedinglab/hh-suite/issues/167

Which link did you download your hhblits database from, and how are they different for the hhblits runs / results? uniclust30_2018_08_hhsuite.tar.gz = 25G, @ http://gwdu111.gwdg.de/~compbiol/uniclust/2018_08/uniclust30_2018_08_hhsuite.tar.gz OR uniclust30_2018_08.tar.gz = 6.6G, @ http://gwdu111.gwdg.de/~compbiol/uniclust/2018_08/uniclust30_2018_08.tar.gz

Since you are running hhblits successfully, I thought you might know :) Could you help please? Thanks!

Log in to answer this question.