Performance of protein-structure predictions with the physics-based UNRES force field in CASP11

Abstract

Summary: Participating as the Cornell-Gdansk group, we have used our physics-based coarsegrained UNited RESidue (UNRES) force field to predict protein structure in the 11th Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction (CASP11). Our methodology involved extensive multiplexed replica exchange simulations of the target proteins with a recently improved UNRES force field to provide better reproductions of the local structures of polypeptide chains. All simulations were started from fully extended polypeptide chains, and no external information was included in the simulation process except for weak restraints on secondary structure to enable us to finish each prediction within the allowed 3-week time window. Because of simplified UNRES representation of polypeptide chains, use of enhanced sampling methods, code optimization and parallelization and sufficient computational resources, we were able to treat, for the first time, all 55 human prediction targets with sizes from 44 to 595 amino acid residues, the average size being 251 residues. Complete structures of six singledomain proteins were predicted accurately, with the highest accuracy being attained for the T0769, for which the CaRMSD was 3.8 Å for 97 residues of the experimental structure. Correct structures were also predicted for 13 domains of multi-domain proteins with accuracy comparable to that of the best template-based modeling methods. With further improvements of the UNRES force field that are now underway, our physics-based coarse-grained approach to protein-structure prediction will eventually reach global prediction capacity and, consequently, reliability in simulating protein structure and dynamics that are important in biochemical processes. Availability and Implementation: Freely available on the web at

Ergodicity and model quality in template‐restrained canonical and temperature/Hamiltonian replica exchange coarse‐grained molecular dynamics simulations of proteins

“…As in the CASP exercises, for the HREMD runs with the TSET2 and WSET3 combinations, we carried out cluster analyses to obtain five families of conformations, from which to select the 5 candidate predictions. The results are shown and compared with the C α ‐RMSDs for the best starting models in Figure d.…”

“…The probabilities of the conformations were calculated at T = 300 K, the conformations were ranked according to decreasing probability, and those conformations were selected for clustering whose summary probability was 0.99. This reduced set was dissected into five clusters to follow our selection procedure of the CASP exercises, in which 5 clusters of conformations are generated to select the five models to be submitted. As in our earlier work, the average conformation (averaged with taking into account the probabilities) was subsequently computed for each cluster and the conformation closest to the average conformation was selected as the representative of the cluster.…”

Prediction of protein structure with the coarse‐grained UNRES force field assisted by small X‐ray scattering data and knowledge‐based information

“…A scheme illustrating the prediction procedure used in this work, which we established during the participation in previous CASP exercises is shown in Figure . As shown in the Figure , the prediction starts with performing multiplexed replica exchange runs with the coarse‐grained UNRES force field .…”

“…As shown in the Figure , the prediction starts with performing multiplexed replica exchange runs with the coarse‐grained UNRES force field . In this work restraints derived from SAXS and server models were imposed; however, it should be noted that this stage can also be carried out in the ab initio mode, with no restraints . The obtained ensemble of conformations is subsequently processed with weighted histogram analysis method (WHAM) and the heat‐capacity profile is calculated to determine the temperature at which the ensemble is analyzed, T a , which is taken as 20 K below the temperature of the major heat‐capacity peak .…”

“…The clusters (and, thereby, the models derived from them later in the process) are ranked by clusters' free energies. A weighted‐average conformation is calculated for each cluster (with weights determined by WHAM) and the closest conformation of the cluster is selected to represent the entire cluster . Each of the coarse‐grained models of the protein under consideration obtained in this way is subsequently converted to all‐atom representation by using the PULCHRA and SCWRL knowledge‐based algorithm for all‐atom backbone and side‐chain reconstruction, respectively, and subjected to final refinement at the all‐atom level with the AMBER 12 force field…”