Protocols for Molecular Modeling with Rosetta3 and RosettaScripts

Abstract

Previously, we published an article providing an overview of the Rosetta suite of biomacromolecular modeling software and a series of step-by-step tutorials [Kaufmann, K. W., et al. (2010) Biochemistry 49, 2987-2998]. The overwhelming positive response to this publication we received motivates us to here share the next iteration of these tutorials that feature de novo folding, comparative modeling, loop construction, protein docking, small molecule docking, and protein design. This updated and expanded set of tutorials is needed, as since 2010 Rosetta has been fully redesigned into an object-oriented protein modeling program Rosetta3. Notable improvements include a substantially improved energy function, an XML-like language termed "RosettaScripts" for flexibly specifying modeling task, new analysis tools, the addition of the TopologyBroker to control conformational sampling, and support for multiple templates in comparative modeling. Rosetta's ability to model systems with symmetric proteins, membrane proteins, noncanonical amino acids, and RNA has also been greatly expanded and improved.

Figure 1

Multitemplate comparative modeling with Rosetta. (A) General workflow of the RosettaCM protocol. (B) Fragment insertion (blue, before insertion; red, after insertion). (C) Recombination of template segments. (D) Fragment insertion and minimization for loop closure. Reprinted with permission from ref (5). Copyright 2013 Elsevier.

Figure 2

Protein–peptide interface prediction using FlexPepDock ab initio. Structure prediction of the Che-Z-derived peptide bound to CheY (PDB entry 2FMF) from two opposite starting orientations converges onto the same final conformation resembling the structure of the native peptide. The left panel is a general view of the CheY receptor (gray; interface residues colored light brown), the two initial, extended peptide conformations (rainbow cartoons), and the final helical peptide conformation (rainbow, transparent cartoon). The right panel is a detailed atomic view of the top FlexPepDock ab initio predictions from two simulations (yellow and orange) and the native peptide conformation (green). Reprinted from ref (15).

Figure 3

Application of RosettaLigand docking of negative allosteric modulator MPEP into a comparative model of the mGlu5 transmembrane domain. The predicted lowest-energy MPEP docking position (cyan) is close to residues demonstrating a change in MPEP modulations upon mutation (yellow to red). Reprinted with permission from ref (108). Copyright 2013 American Society for Pharmacology and Experimental Therapeutics.

Figure 4

Design of protein–ligand interactions for high affinity and selectivity. (A) The design approach involved specifying binding interactions between the protein and ligand followed by design of the binding site. Finally, only designs in which shape complementarity was better than what is seen in native complexes were selected for experimental characterization. (B) Design crystal structure (purple) and computational model (gray) of the protein–ligand complex resulting from design for high affinity and selectivity. The RMSD was 0.54 Å, while the bound form (C) had an RMSD of 0.99 Å. Reprinted with permission from ref (11). Copyright 2013 Macmillan Publishers Ltd.

Figure 5

Stepwise assembly (SWA) structure modeling method for RNA. Illustration of the J2/4 loop from the three-way junction of a TPP-sensing riboswitch (PDB entry 3DV2). (A) Crystallographic conformation of the five-nucleotide loop (colored). (B) Schematic of the three-way junction. (C–F) The loop is built in a stepwise manner, starting from the 3′ end. (G) A directed acyclic graph recursively covers all possible build-up paths. The steps shown in panels C–F are colored magenta. Gray vertices correspond to the starting point with none of the loop nucleotides built. Black vertices are partially built subregions. Red vertices correspond to the ending points with the loop completely built. (H) Energy vs RMSD from the crystal for models generated by SWA (blue points) and by the prior method (FARFAR, red points). The SWA fourth lowest-energy cluster center (purple circle) is within atomic accuracy of the crystallographic model (0.85 Å RMSD). Reprinted with permission from ref (109). Copyright 2011 National Academy of Sciences.