An introductory tutorial on scoring biomolecules using Rosetta can be found here.
REF2015 was developed as beta_nov15 and became the default scorefunction in July 2017. The main changes include optimization of electrostatic parameters, updated torsion parameters, updated bonded parameters, enabling LJ attraction for hydrogens. For more information see beta_nov15_updates and the following paper:
The previous score function, talaris2014, is a slight modification of the talaris2013 energy function. The talaris2013 and talaris2014 energy functions and their corrections were tested in the papers
A full description of the changes this energy function introduces can be found here.
The talaris2014 energy function is suitable for scoring canonical L-amino acids, their D-amino acid mirror images, and some rigid ligands (e.g. metal ions, phosphate, etc.). It can also work with noncanonical alpha-amino acid residues, provided that their params files are set up properly. Backbone conformation terms will ignore beta-amino acids, flexible ligands, nucleic acids, etc.
O'Meara, M. J., Leaver-Fay, A., Tyka, M., Stein, A., Houlihan, K., DiMaio, F., Bradley, P., Kortemme, T., Baker, D., Snoeyink, J., A Combined Covalent-Electrostatic Model of Hydrogen Bonding Improves Structure Prediction with Rosetta. Journal of Chemical Theory and Computation, 2015.
Leaver-Fay, A., O'Meara, M. J., Tyka, M., Jacak, R., Song, Y., Kellogg, E. H., Thompson, J., Davis, I. W., Pache, R. A., Lyskov, S., Gray, J. J., Kortemme, T., Richardson, J. S., Havranek, J. J., Snoeyink, J., Baker, D., Kuhlman, B., Scientific benchmarks for guiding macromolecular energy function improvement. Methods in enzymology, 2013. 523: p. 109.
Rohl, C. A., Strauss, C. EM., Misura, K. MS., Baker, D., Protein structure prediction using Rosetta. Methods in enzymology, 2004. 383: p. 66-93.
Kuhlman, B., Dantas, G., Ireton, G. C., Varani, G., Stoddard, B. L., Baker, D., Design of a novel globular protein fold with atomic-level accuracy. Science, 2003. 302(5649): p. 1364-8.
Kuhlman, B. and D. Baker, Native protein sequences are close to optimal for their structures. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(19): p. 10383-8.
Also, an older presentation about scorefunctions exists.
fa_atr Lennard-Jones attractive between atoms in different residues. Supports canonical and noncanonical residue types. fa_rep Lennard-Jones repulsive between atoms in different residues. Supports canonical and noncanonical residue types. fa_sol Lazaridis-Karplus solvation energy. Supports canonical and noncanonical residue types. fa_intra_rep Lennard-Jones repulsive between atoms in the same residue. Supports canonical and noncanonical residue types. fa_elec Coulombic electrostatic potential with a distance-dependent dielectric. Supports canonical and noncanonical residue types. pro_close Proline ring closure energy and energy of psi angle of preceding residue. Supports D- or L-proline, plus D- or L-oligourea-proline. hbond_sr_bb Backbone-backbone hbonds close in primary sequence. All hydrogen bonding terms support canonical and noncanonical types. hbond_lr_bb Backbone-backbone hbonds distant in primary sequence. hbond_bb_sc Sidechain-backbone hydrogen bond energy. hbond_sc Sidechain-sidechain hydrogen bond energy. dslf_fa13 Disulfide geometry potential. Supports D- and L-cysteine disulfides, plus homocysteine disulfides or disulfides involving beta-3-cysteine. rama Ramachandran preferences. Supports only the 20 canonical alpha-amino acids and their mirror images. omega Omega dihedral in the backbone. A Harmonic constraint on planarity with standard deviation of ~6 deg. Supports alpha-amino acids, beta-amino acids, and oligoureas. In the case of oligoureas, both amide bonds (called "mu" and "omega" in Rosetta) are constarined to planarity. fa_dun Internal energy of sidechain rotamers as derived from Dunbrack's statistics (2010 Rotamer Library used in Talaris2013). Supports any residue type for which a rotamer library is avalable. p_aa_pp Probability of amino acid at Φ/Ψ. Supports only the 20 canonical alpha-amino acids and their mirror images. ref Reference energy for each amino acid. Balances internal energy of amino acid terms. Plays role in design. Supports only the 20 canonical alpha-amino acids and their mirror images. METHOD_WEIGHTS Not an energy term itself, but the parameters for each amino acid used by the ref energy term. A value is provided for each of the 20 canonical alpha-amino acids. The same value is applied for the equivalent mirror-image D-amino acid.
lk_ball Anisotropic contribution to the solvation. Supports arbitrary residue types. lk_ball_iso Same as fa_sol; see below. Supports arbitrary residue types. lk_ball_wtd weighted sum of lk_ball & lk_ball_iso (w1*lk_ball + w2*lk_ball_iso); w2 is negative so that anisotropic contribution(lk_ball) replaces some portion of isotropic contribution (fa_sol=lk_ball_iso). Supports arbitrary residue types. lk_ball_bridge Bonus to solvation coming from bridging waters, measured by overlap of the "balls" from two interacting polar atoms. Supports arbitrary residue types. lk_ball_bridge_uncpl Same as lk_ball_bridge, but the value is uncoupled with dGfree (i.e. constant bonus, whereas lk_ball_bridge is proportional to dGfree values). Supports arbitrary residue types. fa_intra_atr_xover4 Intra-residue LJ attraction, counted for the atom-pairs beyond torsion-relationship. Supports arbitrary residues types. fa_intra_rep_xover4 Intra-residue LJ repulsion, counted for the atom-pairs beyond torsion-relationship. Supports arbitrary residues types. fa_intra_sol_xover4 Intra-residue LK solvation, counted for the atom-pairs beyond torsion-relationship. Supports arbitrary residues types. fa_intra_elec Intra-residue Coulombic interaction, counted for the atom-pairs beyond torsion-relationship. Supports arbitrary residues types. rama_prepro Backbone torsion preference term that takes into account of whether preceding amono acid is Proline or not. Currently supports the 20 canonical alpha-amino acids, their mirror-image D-amino acids, oligoureas, and N-methyl amino acids. Arbitrary new building-blocks can also be supported provided that an N-dimensional mainchain potential can be generated somehow. hxl_tors Sidechain hydroxyl group torsion preference for Ser/Thr/Tyr, supersedes yhh_planarity (that covers L- and D-Tyr only).
Previous versions of Rosetta used the score12 energy function as the default full atom energy function. Many of the energy terms are the same as talaris2013 (though at different weights, and with different parameters), although other terms were also used:
fa_pair Statistics-based pair term, favors salt bridges (replaced by fa_elec in talaris2013). Supported only the 20 canonical alpha-amino acids. fa_plane π-π interaction between aromatic groups, by default = 0. dslf_ss_dst Distance score in current disulfide (replaced by dslf_fa13 in talaris2013). dslf_cs_ang Csangles score in current disulfide (replaced by dslf_fa13 in talaris2013). dslf_ss_dih Dihedral score in current disulfide (replaced by dslf_fa13 in talaris2013). dslf_ca_dih Cα dihedral score in current disulfide (replaced by dslf_fa13 in talaris2013).
The score12 energy function can be used in current Rosetta versions, but the option
-restore_pre_talaris_2013_behavior must be passed.
Overview of Seattle Group energy function optimization project: New beta score function (under active development)
Shapovalov, M.V. and R.L. Dunbrack, A smoothed backbone-dependent rotamer library for proteins derived from adaptive kernel density estimates and regressions. Structure, 2011. 19(6): p. 844-858.
Dunbrack, R.L. and F.E. Cohen, Bayesian statistical analysis of protein side‐chain rotamer preferences. Protein Science, 1997. 6(8): p. 1661-1681.
Lazaridis, T. and M. Karplus, Effective energy function for proteins in solution. Proteins: Structure, Function, and Bioinformatics, 1999. 35(2): p. 133-152.