Fixed Backbone Design Application

This application is based on the RosettaDesign algorithm. All simulations made in design make use of an algorithm for packing amino acid side-chains. To simplify the search, side-chains are only considered in a discrete set of favorable conformations, called rotamers. Rosetta uses Roland Dunbrack's backbone dependent library as well as collection of side-chain conformational libraries assembled by Steven Mayo. Rotamers built with Dunbrack's description have ideal geometries, and the side-chain conformational libraries contain rotamers harvested directly from the PDB with non-ideal bond lengths and angles.

Fixed backbone design strategy

  1. Pick a random sequence to start the simulation or start from a specified sequence. In general either approach gives similar results.
  2. Make a single amino acid substitution or change the conformation (rotamer) of a single amino acid.
  3. Apply the energy function (See the section called "The Energy Function") to determine the new energy of the protein.
  4. Accept or reject the substitution base on the Metropolis criterion: substitutions that raise the energy are accepted at some probability that depends on the local "temperature". The sequence optimization algorithm starts a permissive "temperature" and cools until it reaches "0 degrees".
  5. Repeat. The number of repetitions is empirically derived. For a hundred-residue protein a few hundred thousand rotamer substitutions are attempted.

Input Files

  1. PDB files: The input files for fixed backbone design can be a single pdb or a list of pdb names (the listed pdb files must be avaliable locally)
  2. Resfile: The Resfile specifies which residues will be varied. Please check Resfile Syntax for details.

Command line arguments

  1. Sample Command:
    fixbb.linuxgccrelease @flags > fixbb.log (flags is the flag file) 
  2. Options used in Fixed Backbone Design.
    *General Options:
    -in:file:s my.pdb            Name(s) of single PDB file(s) to process. [FileVector]
    -in:file:l list_of_pdbs      File(s) containing list(s) of PDB files to process. [FileVector]
    -no_optH                     Dont change positions of Hydrogen atoms! (default true, specify false if you want optH) [Boolean]
    -centroid_input              Enable centroid inputs of PDBs. default = 'false'. [Boolean]
    -resfile                     Path to resfile.  Most protocols use only the first and will ignore the rest; it does not track against -s or -l automatically.[FileVector]
    -nstruct                     Number of output structures per use. [Integer]
    *Other Options:
    -ignore_unrecognized_res     Do not abort if unknown residues are found in PDB file; instead, ignore them. default='false'. [Boolean]
    -score:weight                Name of weights file (without extension .wts), default is 'standard'. [String]
    -design_contrast             Output list comparing design sequence to native sequence, default = 'redesign' [File]
    Normal IO and scorefunction flags also apply.

Result Interpretation

There will be 1 or more pdb files generated. The output records the designed protein coordinates and weighted scores. Here are the terms for scores used in Fixed Backbone Design.
fa_atr:                   lennard-jones attractive
fa_rep:                   lennard-jones repulsive
fa_sol:                   lazaridis-karplus solvation energy
fa_intra_rep:             Lennard-jones repulsive between atoms in the same residue
pro_close:                proline ring closure energy
fa_pair:                  statistics based pair term, favors salt bridges
hbond_sr_bb:              backbone-backbone hbonds close in primary sequence
hbond_lr_bb:              backbone-backbone hbonds distant in primary sequence
hbond_bb_sc:              sidechain-backbone hydrogen bond energy
hbond_sc:                 sidechain-sidechain and sidechain-backbone hydrogen bond energy.
dslf_ss_dst:              distance score in current disulfide
dslf_cs_ang:              csangles score in current disulfide
dslf_ss_dih:              dihedral score in current disulfide
dslf_ca_dih:              ca dihedral score in current disulfide
rama:                     ramachandran preferences
omega:                    omega angles
fa_dun:                   internal energy of sidechain rotamers as derived from Dunbrack's statistics.
p_aa_p:                   probability of observing an amino acid, given its phi/psi energy method declaration
ref:                      reference energy for each amino acid
total:                    final score

Generated on Tue Apr 20 03:15:03 2010 for Rosetta Projects by  doxygen 1.5.2