Improvements to Robotics-Inspired Conformational Sampling in Rosetta

By Amelie Stein, Tanja Kortemme. 2013

184 citations recorded by the source corpus.

To accurately predict protein conformations in atomic detail, a computational method must be capable of sampling models sufficiently close to the native structure. All-atom sampling is difficult because of the vast number of possible conformations and extremely rugged energy landscapes. Here, we test three sampling strategies to address these difficulties: conformational diversification, intensification of torsion and omega-angle sampling and parameter annealing. We evaluate these strategies in the context of the robotics-based kinematic closure (KIC) method for local conformational sampling in Rosetta on an established benchmark set of 45 12-residue protein segments without regular secondary structure. We quantify performance as the fraction of sub-Angstrom models generated. While improvements with individual strategies are only modest, the combination of intensification and annealing strategies into a new “next-generation KIC” method yields a four-fold increase over standard KIC in the median percentage of sub-Angstrom models across the dataset. Such improvements enable progress on more difficult problems, as demonstrated on longer segments, several of which could not be accurately remodeled with previous methods. Given its improved sampling capability, next-generation KIC should allow advances in other applications such as local conformational remodeling of multiple segments simultaneously, flexible backbone sequence design, and development of more accurate energy functions.

Subfields: Materials Chemistry, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Materials Chemistry

In Molecular Biology

Citations in the corpus

Accurate structure prediction of CDR H3 loops enabled by a novel structure-based C-terminal constraint (2016, methodology): “On the same set of loops used to benchmark KIC, NGK generates substantially more near-native models (43), which is why it is used as the starting point in this study.”
Protein Loop Modeling Using a New Hybrid Energy Function and Its Application to Modeling in Inaccurate Structural Environments (2014, methodology): “In addition , the results are comparable to those of NGK which carries out extended optimization of environment ( average RMSD of 3 .”
A Web Resource for Standardized Benchmark Datasets, Metrics, and Rosetta Protocols for Macromolecular Modeling and Design (2015, methodology): “(vi) The KIC and NGK protocols do not preserve protein secondary structure, due to probabilistic sampling of φ/ψ torsions from Ramachandran space.”
Blind prediction performance of RosettaAntibody 3.0: Grafting, relaxation, kinematic loop modeling, and full CDR optimization (2014, methodology): “47 KIC randomly perturbs several loop angles and then solves for the remaining six torsions of three “pivot” residues to close the loop using a fast analytical formulation.”
Deep Learning in Protein Structural Modeling and Design (2020, background): “Loops are irregular and sometimes flexible segments, and thus their structures have been difficult to capture experimentally or computationally.(176,177) So far, DL frameworks based on interresidue distance prediction (similar to protein structure prediction),(178) and those based on treating the lo”
Rosetta FunFolDes – A general framework for the computational design of functional proteins (2018, methodology): “…fragment picking Fragment insertion is a core algorithm in Rosetta protocols exploring high degrees of freedom of the polypeptide chain, such as ab initio protein prediction (Simons, Ruczinski, et al., 1999), loop modeling (Stein & Kortemme, 2013), or more recently, FFL (Correia et al., 2014).”
Sphinx: merging knowledge-based and ab initio approaches to improve protein loop prediction (2017, methodology): “This uses the next-generation kinematic modelling (KIC) approach described by Stein and Kortemme (2013) – the full command used is given in the Supplementary Information.”
Design of Protein Multi-specificity Using an Independent Sequence Search Reduces the Barrier to Low Energy Sequences (2015, methodology): “Any chain breaks were closed using kinematic loop closure [42].”
Modeling and docking antibody structures with Rosetta (2017, methodology): “In the KIC loop benchmark 21,42 , loops of 12–17 residues are modeled to near a 1-Å backbone RMSD relative to the native structure—the average human CDR H3 falls within that range, with an average length of 15 residues (under the international Immunogenetics (IMGT) database definition of CDR H3) 43…”
The origin of consistent protein structure refinement from structural averaging. (2015, methodology): “Complementary to this approach are loop and terminus backbone modeling methods (Park & Seok, 2012) (Stein & Kortemme, 2013).”
Essentials of de novo protein design: Methods and applications (2018, methodology): “KIC is a robotics-inspired backbone conformational sampling method that was originally developed for loop prediction 51,56 and that has been recently extended for de novo design of loops and peptides.”
TCRmodel: high resolution modeling of T cell receptors from sequence (2018, methodology): “…of the CDR3 loops (available as an option on the TCRmodel server) is performed using the kinematic closure loop modeling refinement protocol (21), using the recently described REF15 function (22) to perform loop sampling and select a top model among a set of 100 candidate refined loop…”
Algorithms for Protein Design (2016, methodology): “To address this problem, Kortemme and co-workers developed the KIC algorithm [23 ,24], which generates a library of backbone loops for both protein design and protein modeling.”