Documentation for the backrub application

Metadata

This document was edited Dec 4th 2008 by Colin Smith. This application in mini was created by Colin Smith.

References

• Smith CA, Kortemme T. Backrub-like backbone simulation recapitulates natural protein conformational variability and improves mutant side-chain prediction. J Mol Biol. 2008 Jul 18;380(4):742-56. http://dx.doi.org/10.1016/j.jmb.2008.05.023
• Davis IW, Arendall WB 3rd, Richardson DC, Richardson JS. The backrub motion: how protein backbone shrugs when a sidechain dances. Structure. 2006 Feb;14(2):265-74. http://dx.doi.org/10.1016/j.str.2005.10.007
• Betancourt MR. Efficient Monte Carlo trial moves for polypeptide simulations. J Chem Phys. 2005 Nov 1;123(17):174905. http://dx.doi.org/10.1063/1.2102896

Purpose and Algorithm

The backrub algorithm rotates local segments of the protein backbone about an axis defined by the starting and ending atoms of the segment. When combined with side chain sampling, it can be useful for modeling mutations, creating ensembles of protein backbones, and detailed refinement of backbon/side chain conformations.

Differences From Published Rosetta++

One of the primary differences between this implementation and the previous implementation is that backrub is now atom-centric rather than residue-centric. Instead of just CA atoms, any backbone atoms can be used as pivots. For continuity, only CA atoms are enabled as pivots by default.

This changes how the minimum and maximum segment sizes are specified. In the previous implementation, the smallest segment size was 2 residues, which corresponded to rotating the peptide bond (C and N atoms) between the two residues. In this implementation, that would correspond to a segment size of 4 atoms (CA, C, N, and CA). Because of the atom-centric implementation, the smallest possible size is now 3 atoms, which is the default. The previous default largest segment size, 12 residues, corresponds to the new default of 34 atoms. Because only CA atoms are enabled as pivots by default, the possible segments will be identical to those from the previous implementation.

Another difference is that this implementation uses continuous sampling of side chain chi angles instead of fixed rotamers. The angles are chosen according to the Dunbrack rotamer library probabilities. First, a rotamer well is chosen according to its probability of occurrence in the PDB. Second, the individual chi angles are chosen using gaussian distributed random angles with the means and standard deviations from the PDB.

Finally, backbone and side chain sampling are totally decoupled. They use different movers that are randomly chosen at each Monte Carlo step.

Command line arguments

You can run the backrub application with the following flags:

-database                 Database file input search paths

Structure Input

-s                        Name(s) of single PDB file(s) to process
-l                        File(s) containing list(s) of PDB files to process
-ignore_unrecognized_res  Do not abort if unknown residues are found in PDB
                          file; instead, ignore them.

If the PDB file contains a line beginning with FOLD_TREE, then the fold tree specified by that line will be used. The same fold tree will be appended to all resulting structures using the same format. See the documentation for fold trees for more information.

Simulation Options

-nstruct                  Number of times to process each input PDB
-backrub:ntrials          number of Monte Carlo trials to run
-mc_kt                    value of kT for Monte Carlo
-mm_bend_weight           weight of mm_bend bond angle energy term
-cst_fa_weight            weight of the fullatom constraint terms
-cst_fa_file              constraints filename(s) for fullatom. When multiple 
                          files are given a *random* one will be picked.

See the documentation about contraint files for more information.

Backrub Sampling

-pivot_residues           residues for which contiguous stretches can contain
                          segments (internal residue numbers, defaults to all
                          residues)
-pivot_atoms              main chain atoms usable as pivots
-min_atoms                minimum backrub segment size (atoms)
-max_atoms                maximum backrub segment size (atoms)

Side Chain Sampling

-resfile                  resfile filename(s). only the first is used.
-sc_prob                  probability of making a side chain move
-sc_prob_uniform          probability of uniformly sampling chi angles
-initial_pack             force a repack at the beginning regardless of 
                          whether mutations are set in the resfile

Backbone Torsion Angle Sampling

-movemap                  MoveMap file specifying flexible torsions
-sm_prob                  probability of making a small move

For more information about other application flags, please check the options list

Examples

This example simulates loop 6 of triosephosphate isomerase. Note that the resfile uses residue numbering from the PDB. The pivot residues are specified using absolute internal residue numbering.

cat << END_RESFILE > 2YPI.resfile
NATRO
start
3   A NATAA
7   A NATAA
95  A NATAA
96  A NATAA
129 A NATAA
131 A NATAA
134 A NATAA
139 A NATAA
164 A NATAA
165 A NATAA
167 A NATAA
168 A NATAA
170 A NATAA
172 A NATAA
174 A NATAA
177 A NATAA
179 A NATAA
180 A NATAA
183 A NATAA
208 A NATAA
211 A NATAA
216 A NATAA
219 A NATAA
220 A NATAA
223 A NATAA
230 A NATAA
END_RESFILE
backrub -database rosetta3_database -s 2YPI.pdb -ignore_unrecognized_res -resfile 2YPI.resfile
        -pivot_residues 127 128 129 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178

---