Prerequisites

1. NCBI's BLAST+ suite
2. uniref90 seq database prepared for BLAST+
3. dssp. This is included in bin/ but you may need to download and compile the source

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/metsite.git % cd metsite/ansible_installer

Next edit the the config_vars.yml to reflect where you would like metside and any underlying data to be installed. You can choose a version of uniref but we recommend uniref90 for most accurate results.

You can now run ansible as per

% ansible-playbook -i hosts install.yml

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

Installation

> cd src/

> make clean

> make

> make install

Example use

Run checkchains, to check that the chain in your pdb file is present

cp example/1IAR.pdb ./

> ./bin/checkchains_metsite.pl ./example/1IAR.pdb A

Extract fasta seq

> ./bin/fasta_pdbv3.pl ./example/1IAR.pdb ./1IAR.fasta A

run psiblast

> ncbi-blast-2.2.31+/bin/psiblast -query 1IAR.fasta -inclusion_ethresh 0.001 -num_iterations 3 -num_alignments 0 -out_pssm 1IAR.chk -db uniref90.fasta > 1IAR.bls

run chkparse

> ./bin/chkparse 1IAR.chk > 1IAR.mtx

run mtx_pro

> ./bin/mtx_pro 1IAR.mtx 1IAR.pro

Run dssp

> ./bin/dssp 1IAR.pdb 1IAR.dssp > 1IAR.dsspout

Run seed site find. You must use the chain A you tested for in step 1. You need to provide an ion from this list (note the leading underscore) CU|CA|FE|ZN|MG|MN

> ./bin/seedSiteFind 1IAR.pdb _CU 10 1IAR.pro 1IAR.dssp A

run returnNetCut, keeping the atom choice the same, the second arg is a flag which returns the value for the following FPR:

• 0 = 1%
• 1 = 5%
• 2 = 10%
• 3 = 20%

java -cp 'src/org/ucl/conf/:src' returnNetCut CU 1

Take the float returned and use it as part of the input for met_pred run Met Pred, (i,e: 0.182). Lower case metals this time! And provide a path for the data files and chain id

> ./bin/MetPred 1IAR.pdb.DATA cu 0.182 1IAR_MetPred.pdb A ./data/

As an alternatIve to steps 8 and 9 you can run the python script metpred.py which wraps these two steps together. Basically you pass it the options the previous 2 commands need. Note that the classpath must have full canonical paths, and you must provide a location for the metpred binaries

python ./bin/metpred.py /Users/dbuchan/Code/metsite/src/org/ucl/conf/:/Users/dbuchan/Code/metsite/src/ CU 1 1IAR.pdb.DATA 1IAR_MetPred.pdb A ./bin/ ./data/

or the shell script MetPred_wrapper.sh /opt/Code/metsite/bin/MetPred_wrapper.sh /opt/Code/metsite CU 1 1IAR.pdb.DATA 1IAR.pdb A