IntPred is a library for the prediction of protein-protein interface sites from PDB structures. The library can be used to generate features from PDB files, create datasets, train and/or test a learner and generate prediction labels for unlabelled protein structures.
wget https://github.com/ACRMGroup/IntPred/archive/v0.5.tar.gz
tar xvf v0.5.tar.gz
cd IntPred-0.5
./install.pl
source ./setup.sh
The install script assumes you are using RedHat/CentOS/Fedora. See below if you are not.
You need to have sudo permissions or do the install as root.
You are recommended to update your operating system before attempting an install with:
yum update
If you haven't previously used CPAN to install Perl modules, you may need to do the following:
sudo /usr/bin/perl -MCPAN -e shell
Simply accept all the defaults, and when asked for a CPAN mirror, you can select one from http://www.cpan.org/SITES.html - for example ftp://mirror.ox.ac.uk/sites/www.cpan.org/
Then enter
o conf commit
quit
Simply run the install.sh script:
./install.sh
Simply press return to accept all defaults on the initial install. When reinstalling you can skip some of the stages if needed.
The deault install will use the version of perl in /usr/bin/perl. If
you wish to use a different perl install then you should do:
./install.sh /path/to/perl
Now test the install with
./runTests.sh
See below for more details of what happens during the install.
- First you need to set environment variables and add the bin directory to your path:
source ./setup.sh
- IntPred currently only works with files deposited in the PDB and is designed to be able to be run on multiple PDB files in one go. This is done by creating a control file containing, in its simplest form a single line:
pdb : chain : exclchain
For example:
1aut : C : L
If this line is stored in 1autC.dat, the program is then run by typing:
cd $INTPREDBIN
./runIntPred.pl /path/to/1autC.dat > /path/to/1autC.out
Note that you must be in the $INTPREDBIN directory to run the program.
This would predict on chain C of PDB file 1aut ignoring chain
L. exclchain may be blank if no chains are to be ignored.
For full details simply run:
./runIntPred.pl
expat,wget,perl-CPAN,libxml2,libxml2-develandjava-1.8.0-openjdkare installed usingyum. See below if you are using a system other than RedHat/CentOS/Fedora.CPAN is updated and the Module::Build module is installed. Other Perl dependencies are then installed using CPAN including Moose and the latest version of BioPerl.
The trained WEKA model for the predictor is downloaded.
The distribution includes TCNlib and this is unpacked, installed and tested. This downloads a number of other necessary packages.
The only requirement for RedHat-style Linux is for the yum
installation tool. This is used only to install expat, wget,
perl-CPAN, libxml2, libxml2-devel and java-1.8.0-openjdk. If
you are using another Linux version then, from the install.sh script,
comment out the line:
sudo yum -y install expat wget perl-CPAN libxml2 libxml2-devel java-1.8.0-openjdk
Install these packages using your package manager (e.g. apt-get) and
then run the install script.
If you get repeated errors during Perl CPAN installs along the lines of
CPAN::Meta::Requirements not available at ...
then you may need to install this module manually. Download and install with:
cd /var/tmp
wget http://www.cpan.org/authors/id/D/DA/DAGOLDEN/CPAN-Meta-Requirements-2.140.tar.gz
tar xvf CPAN-Meta-Requirements-2.140.tar.gz
cd CPAN-Meta-Requirements-2.140
perl Makefile.PL
make all
make test && sudo make install
then re-run install.sh
Note that all tests on TCNlib should pass (except the ones that
require PyMol if you don't have that installed). If they don't then
something has probably gone wrong with the Perl module installs. Check
for error messages and then use CPAN to install the missing modules;
if there have been problems with CPAN-Meta-Requirements then you
will probably have to do this one at a time checking the dependencies
manually.