Rosetta 3.5
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Pages
Documentation for the loop modeling application


This document was edited Nov 1st 2008 by Yi Liu. This application in Rosetta3 was created by Michael Tyka, et al. Most of the contents are still based on the Rosetta++ loop tutorial written by Chu Wang 2007.

Example runs

This file only contents loop relax example at this moment. See rosetta/rosetta_tests/integration/tests/loop_modeling for an example loop relax run and input files.


Purpose and Algorithm

This protocol was developed to be combined with Rosetta full atom structure refinement (relax mode) to streamline the task of comparative modeling. Briefly, it identifies structural variable regions from an ensemble of refined models by Rosetta fullatom refinement protocol and then models these loops using a protocol similar to the pose-based protocol.

Input Files

Command Line Options

You can run loop modeling with the following flags:

-in::file::s inputs/4fxn.start_0001.pdb
-loops::loop_file inputs/4fxn.loop_file
-loops::frag_sizes 9 3 1
-loops::frag_files  inputs/cc4fxn_09_05.200_v1_3.gz inputs/cc4fxn_03_05.200_v1_3.gz none

Other combinable flags:

-loops::remodel [perturb_ccd|perturb_alc|quick_ccd|old_looprelax]     Centroid remodelling
-loops::refine [refine_ccd|refine_alc]                                Fullatom refinement
-loops::relax [fullrelax|shortrelax|fastrelax]                        Fullatom relax

For more information about loop modeling flags, please check the full options list

loopmodel MPI

The loopmodel executable has a separate MPI implementation from the JD2 implementation that serves most of Rosetta. As normal, just compile Rosetta in mpi (add extras=mpi to the scons command line when compiling) to activate MPI. When you run loopmodel.mpi.***, it will expect that you have precreated output directories for it, named ./output_#, where # is the zero-indexed processor rank (and you need one for each processor). So, for a 8-processor MPI job, created output_0, output_1, ..., output_7.