Rosetta 3.1 Release Manual

Fixed Backbone ddG Prediction

This application estimates the impact of a sequence change on a protein's stability. It can perform single mutations as well as multiple mutations. There are a variety of options that can be used to refine the structure (this application is still under developement), but only the fixed backbone monomeric protocol will be introduced and described here. The ddg application is in rosetta_source/src/apps/public/ the name of the application is fix_bb_monomer_ddg.cc

Input Files

the input files are pdb files. Sidechain minimization is recommended for the input structures. The second input file specifies mutations to make. Two formats are currently accepted:

resfiles: option to specify resfile: -resfile The resfile specifies which residues to mutate and what to mutate them into. This will produce single mutations for each of the amino acids specified. Please look at Resfile Syntax for details. If, for example, ALLAA is specified at a certain position, the position will be mutated into the 19 other amino acids and the ddg for each of these 19 mutations will be calculated. EX 1 and EX 2 keywords have no effect on how the structures are refined.

"mutfiles": option to specify mutfile: -ddg:mut_file a file for specifying mutations that supports specification of multiple mutations. The format is as follows: header: the total keyword specifies how many mutations to expect. If the number of mutations listed do not match up with the total number declared, no calculations will be performed. content: each mutation is specified in the following manner: the number of mutations to be made is listed, and on the next line is the wild-type residue, the position of the mutation, and the mutated residue desired. And example of a correctly specified mutfile follows. For this example, we want to specify 3 mutations total. one calculation is a double mutant with an Alanine at position 167 to Phenylalanine and a Thr at position 67 to Ala. The second calculation is a single mutation.

total 3
2
A 167 F
T 67 A
1
I 16 A

Command Lines and Options

• Options used by the fix_bb_monomer_ddg application:

  *inputs:
  ----------
  -in:file:s
  	specifies starting structures. accepts only pdb format
  -packing:resfile
  	specifies which mutations to make. only supports single mutations
  -ddg::mut_file
  	alternate format to specify mutations. supports single and multiple mutations.
  
  *outputs:
  ----------
  -ddg::dump_pdbs
  	boolean. defaults to true. specify whether or not to dump repacked wild-type and mutant pdbs. Naming of output pdbs as follows:
  		 wild-type refined decoys
  		 	-repacked_round_X.pdb where X is the iteration number
  		 mutant refined decoys
  		 	-mut_<wild-type_residue>_<position>_<mutant-residue>_round_X.pdb
  -ddg::out
  	string. defaults to ddg_predictions.out. This is the file where all predicted ddgs are written, along with the breakdown of the score-components that contribute to the prediction.
  -ddg::debug_output
  	boolean. defaults to false. specify whether or not to write a whole bunch of debug statements to standard out
  -ddg::output_silent
  	boolean. no default. specifies whether to dump pdbs as silent files. Silent file naming convention should be self-explanatory.
  
  *options
  ----------
  -ddg::weight_file
  	string. defaults to ddg.wts. specifies the weight-files to be used in calculations
  -ddg::iterations
  	Real. defaults to 20. specifies the number of iterations of refinement. This is the number of structures produced by applying refinement to the starting structure.
  -ddg::opt_radius
  	Real. defaults to 8. if local-optimization is preferred over global refinement of the structure,this option specifies how many angstrom from the mutation site a
  residue can be in order to be considered a neighbor.
  -ddg::local_opt_only
  	Boolean. defaults to false. If this is set to true, local optimization of the mutation's neighborhood will be performed instead of global optimization.
  Also, output of silent files will be default behavior.
  -ddg::mean
  	Boolean. no default. specifies whether the mean energy of the ensemble is to be used as the predicted free energy of folding.
  -ddg::min
  	Boolean. no default. specifies whether the minimum energy of the ensemble is to be used as the predicted free energy of folding.
  -ddg::min_cst
  	Boolean. defaults to true. This option is still currently under developement and is not recommended for usage. apologies for the inconvenience
  -ddg::sc_min_only
  	Boolean. defaults to true. Another option that is still currently under developement and is not recommended. again, apologies for the inconvenice to users.
  

• Sample Command Lines
  The following command is the recommended command-line options to achieve the best performance, and has been shown to produce a correlation of R = 0.64 on a set of > 1000 mutations. (capitalized text between "<" and ">" indicates user input)

  1. ./fix_bb_monomer_ddg.linuxgccrelease -in:file:s <PDB file> -ddg::weight_file ddg.wts -ddg::iterations 20
     -ddg::dump_pdbs true -ddg::mut_file <mutation_file.mutfile> -database <PATH TO ROSETTA_DATABASE> -ignore_unrecognized_re
     -ddg::local_opt_only false -ddg::min_cst false -ddg::mean true -ddg::min false -ddg::sc_min_only false  -in:file:fullatom
     

---

© Copyright Rosetta Commons Member Institutions. For more information, see http://www.rosettacommons.org.