The Rosetta software suite for macromolecular modeling, docking, and design is widely used in pharmaceutical, industrial, academic, non-profit, and government laboratories. Despite its broad modeling capabilities, Rosetta remains consistently among leading software suites when compared to other methods created for highly specialized protein modeling and design tasks. Developed for over two decades by a global community of over 60 laboratories, Rosetta has undergone multiple refactorings, and now comprises over three million lines of code. Here we discuss methods developed in the last five years in Rosetta, involving the latest protocols for structure prediction; protein-protein and protein-small molecule docking; protein structure and interface design; loop modeling; the incorporation of various types of experimental data; modeling of peptides, antibodies and proteins in the immune system, nucleic acids, non-standard chemistries, carbohydrates, and membrane proteins. We briefly discuss improvements to the energy function, user interfaces, and usability of the software. Rosetta is available at www.rosettacommons.org.

We present the results for CAPRI Round 30, the first joint CASP-CAPRI experiment, which brought together experts from the protein structure prediction and protein-protein docking communities. The Round comprised 25 targets from amongst those submitted for the CASP11 prediction experiment of 2014. The targets included mostly homodimers, a few homotetramers, and two heterodimers, and comprised protein chains that could readily be modeled using templates from the Protein Data Bank. On average 24 CAPRI groups and 7 CASP groups submitted docking predictions for each target, and 12 CAPRI groups per target participated in the CAPRI scoring experiment. In total more than 9500 models were assessed against the 3D structures of the corresponding target complexes. Results show that the prediction of homodimer assemblies by homology modeling techniques and docking calculations is quite successful for targets featuring large enough subunit interfaces to represent stable associations. Targets with ambiguous or inaccurate oligomeric state assignments, often featuring crystal contact-sized interfaces, represented a confounding factor. For those, a much poorer prediction performance was achieved, while nonetheless often providing helpful clues on the correct oligomeric state of the protein. The prediction performance was very poor for genuine tetrameric targets, where the inaccuracy of the homology-built subunit models and the smaller pair-wise interfaces severely limited the ability to derive the correct assembly mode. Our analysis also shows that docking procedures tend to perform better than standard homology modeling techniques and that highly accurate models of the protein components are not always required to identify their association modes with acceptable accuracy.

We describe Rosetta-based computational protocols for predicting the three-dimensional structure of an antibody from sequence (RosettaAntibody) and then docking the antibody to protein antigens (SnugDock). Antibody modeling leverages canonical loop conformations to graft large segments from experimentally-determined structures as well as (1) energetic calculations to minimize loops, (2) docking methodology to refine the VL–VH relative orientation, and (3) de novo prediction of the elusive complementarity determining region (CDR) H3 loop. To alleviate model uncertainty, antibody–antigen docking resamples CDR loop conformations and can use multiple models to represent an ensemble of conformations for the antibody, the antigen or both. These protocols can be run fully-automated via the ROSIE web server (http://rosie.rosettacommons.org/) or manually on a computer with user control of individual steps. For best results, the protocol requires roughly 1,000 CPU-hours for antibody modeling and 250 CPU-hours for antibody–antigen docking. Tasks can be completed in under a day by using public supercomputers.

Supplementary data are available at Bioinformatics online.

The branched chain amino acids (BCAA) valine, leucine and isoleucine have been implicated in a number of diseases including obesity, insulin resistance, and type 2 diabetes mellitus, although the mechanisms are still poorly understood. Adipose tissue plays an important role in BCAA homeostasis by actively metabolizing circulating BCAA. In this work, we have investigated the link between BCAA catabolism and fatty acid synthesis in 3T3-L1 adipocytes using parallel 13C-labeling experiments, mass spectrometry and model-based isotopomer data analysis. Specifically, we performed parallel labeling experiments with four fully 13C-labeled tracers, [U-13C]valine, [U-13C]leucine, [U-13C]isoleucine and [U-13C]glutamine. We measured mass isotopomer distributions of fatty acids and intracellular metabolites by GC-MS and analyzed the data using the isotopomer spectral analysis (ISA) framework. We demonstrate that 3T3-L1 adipocytes accumulate significant amounts of even chain length (C14:0, C16:0 and C18:0) and odd chain length (C15:0 and C17:0) fatty acids under standard cell culture conditions. Using a novel GC-MS method, we demonstrate that propionyl-CoA acts as the primer on fatty acid synthase for the production of odd chain fatty acids. BCAA contributed significantly to the production of all fatty acids. Leucine and isoleucine contributed at least 25% to lipogenic acetyl-CoA pool, and valine and isoleucine contributed 100% to lipogenic propionyl-CoA pool. Our results further suggest that low activity of methylmalonyl-CoA mutase and mass action kinetics of propionyl-CoA on fatty acid synthase result in high rates of odd chain fatty acid synthesis in 3T3-L1 cells. Overall, this work provides important new insights into the connection between BCAA catabolism and fatty acid synthesis in adipocytes and underscores the high capacity of adipocytes for metabolizing BCAA.

Antibody Modeling Assessment II (AMA-II) provided an opportunity to benchmark RosettaAntibody on a set of eleven unpublished antibody structures. RosettaAntibody produced accurate, physically realistic models, with all framework regions and 42 of the 55 non-H3 CDR loops predicted to under an Ångström. The performance is notable when modeling H3 on a homology framework, where RosettaAntibody produced the best model among all participants for four of the eleven targets, two of which were predicted with sub-Ångström accuracy. To improve RosettaAntibody, we pursued the causes of model errors. The most common limitation was template unavailability, underscoring the need for more antibody structures and/or better de novo loop methods. In some cases, better templates could have been found by considering residues outside of the CDRs. De novo CDR H3 modeling remains challenging at long loop lengths, but constraining the C-terminal end of H3 to a kinked conformation allows near-native conformations to be sampled more frequently. We also found that incorrect VL–VH orientations caused models with low H3 RMSDs to score poorly, suggesting that correct VL–VH orientations will improve discrimination between near-native and incorrect conformations. These observations will guide the future development of RosettaAntibody.

Antibodies are important immune molecules with high commercial value and therapeutic interest because of their ability to bind diverse antigens. Computational prediction of antibody structure can quickly reveal valuable information about the nature of these antigen-binding interactions, but only if the models are of sufficient quality. To achieve high model quality during complementaritydetermining region (CDR) structural prediction, one must account for the V L -V H orientation. We developed a novel four-metric V L -V H orientation coordinate frame. Additionally, we extended the CDR grafting protocol in RosettaAntibody with a new method that diversifies V L -V H orientation by using 10 V L -V H orientation templates rather than a single one. We tested the multiple-template grafting protocol on two datasets of known antibody crystal structures. During the template-grafting phase, the new protocol improved the fraction of accurate V L -V H orientation predictions from only 26% (12/46) to 72% (33/46) of targets. After the full RosettaAntibody protocol, including CDR H3 remodeling and V L -V H re-orientation, the new protocol produced more candidate structures with accurate V L -V H orientation than the standard protocol in 43/46 targets (93%). The improved ability to predict V L -V H orientation will bolster predictions of other parts of the paratope, including the conformation of CDR H3, a grand challenge of antibody homology modeling.

We describe Rosetta-based computational protocols for predicting the three-dimensional structure of an antibody from sequence and then docking the antibody-antigen complexes. Antibody modeling leverages canonical loop conformations to graft large segments from experimentally-determined structures as well as (1) energetic calculations to minimize loops, (2) docking methodology to refine the V L -V H relative orientation, and (3) de novo prediction of the elusive CDR H3 loop. To alleviate model uncertainty, antibody-antigen docking resamples CDR loop conformations and can use multiple models to represent an ensemble of conformations for the antibody, the antigen or both. These protocols can be run fullyautomated via the ROSIE web server or manually on a computer with user control of individual steps. For best results, the protocol requires roughly 2,500 CPU-hours for antibody modeling and 250 CPU-hours for antibody-antigen docking. Both tasks can be completed in under a day by using public supercomputers.