buildModel.pl

Usage

usage:   buildModel.pl [options] [fastaFile]
options:               [-models value]
                       [-maxloop value]
                       [-verbose]
                       [-nocenter]
                       [-stats file]
                       [-pdbdir name]

Show source

Description

This script is used for template-based structure prediction. It takes an alignment file in FASTA file as input, finds the PDB template, builds a homology model and adds missing (loop) residues via Modeller.

A typical input file is expected to look as follows:

>TARGET 
RPRCFFDIAINNQPAGRVVFELFSDVCPKTCENFRCLCTGEKGTGKSTQKPLHYKSCLFHRVVKDFMVQGGDFSEGNGRGGESIYGGFFEDESFAVKHNAAFLLSMANRGKDTNGSQFFITTKPTPHLDGHHVVFGQVISGQEVVREIENQKTDAASKPFAEVRILSCGELIP   
>1IHG_A ::4e-96::61::Chain A, Bovine Cyclophilin 40, Monoclinic Form
NPRVFFDVDIGGERVGRIVLELFADIVPKTAENFRALCTGEKGIGPTTGKPLHFKGCPFHRIIKKFMIQGGDFSNQNGTGGESIYGEKFEDENFHYKHDKEGLLSMANAGSNTNGSQFFITTVPTPHLDGKHVVFGQVIKGMGVAKILENVEVK-GEKPAKLCVIAECGEL..

Note the PDB code and chain ID in the third line. A properly formatted file can be generated with the psiblast.pl tool from PSI-Blast output.

For this tool to work, it is necessary to have a local copy of the PDB (or at least a directory where the structures for the needed templates are present) either at the location pointed to by the environment variable PDBDIR or given through the option -pdbdir.

Furthermore, Modeller needs to be installed and available. The script will try to guess the location of the Modeller script that can be called with a Modeller input file, but it is better to provide the location through the environment variable MODELLEREXEC.

Options

-help 

usage information