Ansible Installation

First ensure that ansible is installed on your system, then clone the github repo

% pip install ansible
% git clone https://github.com/psipred/hspred.git
% cd ./hspred/ansible_installer

Next edit the the config_vars.yml to reflect where you would like hspred.

You can now run ansible as per:

% ansible-playbook -i hosts install.yml

You can edit the hosts file to install hspred on one or more machines.

Prerequsites

1. Install perl Expect.pm
2. clean.ex and hbplus are provided in the bin/ dir if you find they will not run you can get the source files from the EBI. Clean is part of the procheck suite (https://www.ebi.ac.uk/thornton-srv/software/PROCHECK/) and hbplus can be downloaded via (https://www.ebi.ac.uk/thornton-srv/software/HBPLUS/install.html). You need the resdefs.dat file in the data dir
3. svm_classify from svm_light may need recompiled (in data/)

Example use

Check that the chains you want to analyse are present in the pdb file

> cp ./examples/1IAR.pdb ./

> ./bin/checkchains.pl 1IAR.pdb AB ./data/charmm19_ha.aaa

If it is happy your chains are present proceed to the next step

Run cleanexpect, pass the location of the resdefs.dat

> ./bin/clean_expect.pl 1IAR.pdb ./data/resdefs.dat

This will produce a file 1IAR.new

Run hbplus over the cleaned pdb file

> ./bin/hbplus 1IAR.new 1IAR.pdb

This will produce a file 1IAR.hb2

Finally you can run the predictor, hspred

> ./bin/hs-pred_v0_1.pl 1IAR A B ./data ./

Args are the pdb code, the chains for the 2 domains to be analyses a path to the data files and a final temp path for the outputs. The pdb and hb2 files must be in the temp dir provided.

You should find a range of new ouptput files

If you want to visualise the predictions annotated you can run splitpdb. This will output a pdb file for each chain with the ATOM temperature set to indicate the hotspot predictions.

> ./bin/split_pdb.pl 1IAR A B ./

ARGS: pdb code, chain IDs and the working/temp directory