Protein Structure Refinement through Structure Selection and Averaging from Molecular Dynamics Ensembles

Protein Retrieval via Integrative Molecular Ensembles (PRIME) through Extended Similarity Indices

Molecular dynamics (MD) simulations are ideally suited to describe conformational ensembles of biomolecules such as proteins and nucleic acids. Microsecond-long simulations are now routine, facilitated by the emergence of graphical processing units. Clustering, which groups objects based on structural similarity, is typically used to process ensembles, leading to different states, their populations, and the identification of representative structures. A popular pipeline combines hierarchical clustering for clustering and selecting the cluster centroid as representative of the cluster. Here, we propose to improve on this approach, by developing a module-Protein Retrieval via Integrative Molecular Ensembles (PRIME), that consists of tools to improve the prediction of the representative in the most populated cluster using extended continuous similarity. PRIME is integrated with our Molecular Dynamics Analysis with N-ary Clustering Ensembles (MDANCE) package and can be used as a postprocessing tool for arbitrary clustering algorithms, compatible with several MD suites. PRIME predictions produced structures that when aligned to the experimental structure were better superposed (lower RMSD). A further benefit of PRIME is its linear scaling─rather than the traditional O(N2) traditionally associated with comparisons of elements in a set.

Making Use of Averaging Methods in MODELLER for Protein Structure Prediction

Recent advances in protein structure prediction, driven by AlphaFold 2 and machine learning, demonstrate proficiency in static structures but encounter challenges in capturing essential dynamic features crucial for understanding biological function. In this context, homology-based modeling emerges as a cost-effective and computationally efficient alternative. The MODELLER (version 10.5, accessed on 30 November 2023) algorithm can be harnessed for this purpose since it computes intermediate models during simulated annealing, enabling the exploration of attainable configurational states and energies while minimizing its objective function. There have been a few attempts to date to improve the models generated by its algorithm, and in particular, there is no literature regarding the implementation of an averaging procedure involving the intermediate models in the MODELLER algorithm. In this study, we examined MODELLER's output using 225 target-template pairs, extracting the best representatives of intermediate models. Applying an averaging procedure to the selected intermediate structures based on statistical potentials, we aimed to determine: (1) whether averaging improves the quality of structural models during the building phase; (2) if ranking by statistical potentials reliably selects the best models, leading to improved final model quality; (3) whether using a single template versus multiple templates affects the averaging approach; (4) whether the "ensemble" nature of the MODELLER building phase can be harnessed to capture low-energy conformations in holo structures modeling. Our findings indicate that while improvements typically fall short of a few decimal points in the model evaluation metric, a notable fraction of configurations exhibit slightly higher similarity to the native structure than MODELLER's proposed final model. The averaging-building procedure proves particularly beneficial in (1) regions of low sequence identity between the target and template(s), the most challenging aspect of homology modeling; (2) holo protein conformations generation, an area in which MODELLER and related tools usually fall short of the expected performance.

Computational approaches to investigate fluoride binding, selectivity and transport across the membrane

The use of molecular dynamics (MD) simulations to study biomolecular systems has proven reliable in elucidating atomic-level details of structure and function. In this chapter, MD simulations were used to uncover new insights into two phylogenetically unrelated bacterial fluoride (F-) exporters: the CLCF F-/H+ antiporter and the Fluc F- channel. The CLCF antiporter, a member of the broader CLC family, has previously revealed unique stoichiometry, anion-coordinating residues, and the absence of an internal glutamate crucial for proton import in the CLCs. Through MD simulations enhanced with umbrella sampling, we provide insights into the energetics and mechanism of the CLCF transport process, including its selectivity for F- over HF. In contrast, the Fluc F- channel presents a novel architecture as a dual topology dimer, featuring two pores for F- export and a central non-transported sodium ion. Using computational electrophysiology, we simulate the electrochemical gradient necessary for F- export in Fluc and reveal details about the coordination and hydration of both F- and the central sodium ion. The procedures described here delineate the specifics of these advanced techniques and can also be adapted to investigate other membrane protein systems.

Representation of Protein Dynamics Disentangled by Time-Structure-Based Prior

Representation learning (RL) is a universal technique for deriving low-dimensional disentangled representations from high-dimensional observations, aiding in a multitude of downstream tasks. RL has been extensively applied to various data types, including images and natural language. Here, we analyze molecular dynamics (MD) simulation data of biomolecules in terms of RL. Currently, state-of-the-art RL techniques, mainly motivated by the variational principle, try to capture slow motions in the representation (latent) space. Here, we propose two methods based on an alternative perspective on the disentanglement in the latent space. By disentanglement, we here mean the separation of underlying factors in the simulation data, aiding in detecting physically important coordinates for conformational transitions. The proposed methods introduce a simple prior that imposes temporal constraints in the latent space, serving as a regularization term to facilitate the capture of disentangled representations of dynamics. Comparison with other methods via the analysis of MD simulation trajectories for alanine dipeptide and chignolin validates that the proposed methods construct Markov state models (MSMs) whose implied time scales are comparable to those of the state-of-the-art methods. Using a measure based on total variation, we quantitatively evaluated that the proposed methods successfully disentangle physically important coordinates, aiding the interpretation of folding/unfolding transitions of chignolin. Overall, our methods provide good representations of complex biomolecular dynamics for downstream tasks, allowing for better interpretations of the conformational transitions.

When MELD Meets GaMD: Accelerating Biomolecular Landscape Exploration

We introduce Gaussian accelerated MELD (GaMELD) as a new method for exploring the energy landscape of biomolecules. GaMELD combines the strengths of Gaussian accelerated molecular dynamics (GaMD) and modeling employing limited data (MELD) to navigate complex energy landscapes. MELD uses replica-exchange molecular simulations to integrate limited and uncertain data into simulations via Bayesian inference. MELD has been successfully applied to problems of structure prediction like protein folding and complex structure prediction. However, the computational cost for MELD simulations has limited its broader applicability. The synergy of GaMD and MELD surmounts this limitation efficiently sampling the energy landscape at a lower computational cost (reducing the computational cost by a factor of 2 to six). Effectively, GaMD is used to shift energy distributions along replicas to increase the overlap in energy distributions across replicas, facilitating a random walk in replica space. We tested GaMELD on a benchmark set of 12 small proteins that have been previously studied through MELD and conventional MD. GaMELD consistently achieves accurate predictions with fewer replicas. By increasing the efficacy of replica exchange, GaMELD effectively accelerates convergence in the conformational space, enabling improved sampling across a diverse set of systems.

Mechanistic Insight into the Suppression of Polyglutamine Aggregation by SRCP1

Protein aggregation is a hallmark of the polyglutamine diseases. One potential treatment for these diseases is suppression of polyglutamine aggregation. Previous work identified the cellular slime mold Dictyostelium discoideum as being naturally resistant to polyglutamine aggregation. Further work identified serine-rich chaperone protein 1 (SRCP1) as a protein that is both necessary in Dictyostelium and sufficient in human cells to suppress polyglutamine aggregation. Therefore, understanding how SRCP1 suppresses aggregation may be useful for developing therapeutics for the polyglutamine diseases. Here we utilized a de novo protein modeling approach to generate predictions of SRCP1's structure. Using our best-fit model, we generated mutants that were predicted to alter the stability of SRCP1 and tested these mutants' stability in cells. Using these data, we identified top models of SRCP1's structure that are consistent with the C-terminal region of SRCP1 forming a β-hairpin with a highly dynamic N-terminal region. We next generated a series of peptides that mimic the predicted β-hairpin and validated that they inhibit aggregation of a polyglutamine-expanded mutant huntingtin exon 1 fragment in vitro. To further assess mechanistic details of how SRCP1 inhibits polyglutamine aggregation, we utilized biochemical assays to determine that SRCP1 inhibits secondary nucleation in a manner dependent upon the regions flanking the polyglutamine tract. Finally, to determine if SRCP1 more could generally suppress protein aggregation, we confirmed that it was sufficient to inhibit aggregation of polyglutamine-expanded ataxin-3. Together these studies provide details into the structural and mechanistic basis of the inhibition of protein aggregation by SRCP1.

Using Local Protein Model Quality Estimates to Guide a Molecular Dynamics-Based Refinement Strategy

The refinement of predicted 3D models aims to bring them closer to the native structure by fixing errors including unusual bonds and torsion angles and irregular hydrogen bonding patterns. Refinement approaches based on molecular dynamics (MD) simulations using different types of restraints have performed well since CASP10. ReFOLD, developed by the McGuffin group, was one of the many MD-based refinement approaches, which were tested in CASP 12. When the performance of the ReFOLD method in CASP12 was evaluated, it was observed that ReFOLD suffered from the absence of a reliable guidance mechanism to reach consistent improvement for the quality of predicted 3D models, particularly in the case of template-based modelling (TBM) targets. Therefore, here we propose to utilize the local quality assessment score produced by ModFOLD6 to guide the MD-based refinement approach to further increase the accuracy of the predicted 3D models. The relative performance of the new local quality assessment guided MD-based refinement protocol and the original MD-based protocol ReFOLD are compared utilizing many different official scoring methods. By using the per-residue accuracy (or local quality) score to guide the refinement process, we are able to prevent the refined models from undesired structural deviations, thereby leading to more consistent improvements. This chapter will include a detailed analysis of the performance of the local quality assessment guided MD-based protocol versus that deployed in the original ReFOLD method.

Holo Protein Conformation Generation from Apo Structures by Ligand Binding Site Refinement

An important part in structure-based drug design is the selection of an appropriate protein structure. It has been revealed that a holo protein structure that contains a well-defined binding site is a much better choice than an apo structure in structure-based drug discovery. Therefore, it is valuable to obtain a holo-like protein conformation from apo structures in the case where no holo structure is available. Meeting the need, we present a robust approach to generate reliable holo-like structures from apo structures by ligand binding site refinement with restraints derived from holo templates with low homology. Our method was tested on a test set of 32 proteins from the DUD-E data set and compared with other approaches. It was shown that our method successfully refined the apo structures toward the corresponding holo conformations for 23 of 32 proteins, reducing the average all-heavy-atom RMSD of binding site residues by 0.48 Å. In addition, when evaluated against all the holo structures in the protein data bank, our method can improve the binding site RMSD for 14 of 19 cases that experience significant conformational changes. Furthermore, our refined structures also demonstrate their advantages over the apo structures in ligand binding mode predictions by both rigid docking and flexible docking and in virtual screening on the database of active and decoy ligands from the DUD-E. These results indicate that our method is effective in recovering holo-like conformations and will be valuable in structure-based drug discovery.

Physics-based protein structure refinement in the era of artificial intelligence

Protein structure refinement is the last step in protein structure prediction pipelines. Physics-based refinement via molecular dynamics (MD) simulations has made significant progress during recent years. During CASP14, we tested a new refinement protocol based on an improved sampling strategy via MD simulations. MD simulations were carried out at an elevated temperature (360 K). An optimized use of biasing restraints and the use of multiple starting models led to enhanced sampling. The new protocol generally improved the model quality. In comparison with our previous protocols, the CASP14 protocol showed clear improvements. Our approach was successful with most initial models, many based on deep learning methods. However, we found that our approach was not able to refine machine-learning models from the AlphaFold2 group, often decreasing already high initial qualities. To better understand the role of refinement given new types of models based on machine-learning, a detailed analysis via MD simulations and Markov state modeling is presented here. We continue to find that MD-based refinement has the potential to improve AI predictions. We also identified several practical issues that make it difficult to realize that potential. Increasingly important is the consideration of inter-domain and oligomeric contacts in simulations; the presence of large kinetic barriers in refinement pathways also continues to present challenges. Finally, we provide a perspective on how physics-based refinement could continue to play a role in the future for improving initial predictions based on machine learning-based methods.

ReFOLD3: refinement of 3D protein models with gradual restraints based on predicted local quality and residue contacts

ReFOLD3 is unique in its application of gradual restraints, calculated from local model quality estimates and contact predictions, which are used to guide the refinement of theoretical 3D protein models towards the native structures. ReFOLD3 achieves improved performance by using an iterative refinement protocol to fix incorrect residue contacts and local errors, including unusual bonds and angles, which are identified in the submitted models by our leading ModFOLD8 model quality assessment method. Following refinement, the likely resulting improvements to the submitted models are recognized by ModFOLD8, which produces both global and local quality estimates. During the CASP14 prediction season (May-Aug 2020), we used the ReFOLD3 protocol to refine hundreds of 3D models, for both the refinement and the main tertiary structure prediction categories. Our group improved the global and local quality scores for numerous starting models in the refinement category, where we ranked in the top 10 according to the official assessment. The ReFOLD3 protocol was also used for the refinement of the SARS-CoV-2 targets as a part of the CASP Commons COVID-19 initiative, and we provided a significant number of the top 10 models. The ReFOLD3 web server is freely available at https://www.reading.ac.uk/bioinf/ReFOLD/.

Improved Sampling Strategies for Protein Model Refinement Based on Molecular Dynamics Simulation

Protein structures provide valuable information for understanding biological processes. Protein structures can be determined by experimental methods such as X-ray crystallography, nuclear magnetic resonance spectroscopy, or cryogenic electron microscopy. As an alternative, in silico methods can be used to predict protein structures. These methods utilize protein structure databases for structure prediction via template-based modeling or for training machine-learning models to generate predictions. Structure prediction for proteins distant from proteins with known structures often results in lower accuracy with respect to the true physiological structures. Physics-based protein model refinement methods can be applied to improve model accuracy in the predicted models. Refinement methods rely on conformational sampling around the predicted structures, and if structures closer to the native states are sampled, improvements in the model quality become possible. Molecular dynamics simulations have been especially successful for improving model qualities but although consistent refinement can be achieved, the improvements in model qualities are still moderate. To extend the refinement performance of a simulation-based protocol, we explored new schemes that focus on optimized use of biasing functions and the application of increased simulation temperatures. In addition, we tested the use of alternative initial models so that the simulations can explore the conformational space more broadly. Based on the insights of this analysis, we are proposing a new refinement protocol that significantly outperformed previous state-of-the-art molecular dynamics simulation-based protocols in the benchmark tests described here.

Probabilistic divergence of a template-based modelling methodology from the ideal protocol

Protein structural information is essential for the detailed mapping of a functional protein network. For a higher modelling accuracy and quicker implementation, template-based algorithms have been extensively deployed and redefined. The methods only assess the predicted structure against its native state/template and do not estimate the accuracy for each modelling step. A divergence measure is therefore postulated to estimate the modelling accuracy against its theoretical optimal benchmark. By freezing the domain boundaries, the divergence measures are predicted for the most crucial steps of a modelling algorithm. To precisely refine the score using weighting constants, big data analysis could further be deployed.

High-accuracy protein structures by combining machine-learning with physics-based refinement

Protein structure prediction has long been available as an alternative to experimental structure determination, especially via homology modeling based on templates from related sequences. Recently, models based on distance restraints from coevolutionary analysis via machine learning to have significantly expanded the ability to predict structures for sequences without templates. One such method, AlphaFold, also performs well on sequences where templates are available but without using such information directly. Here we show that combining machine-learning based models from AlphaFold with state-of-the-art physics-based refinement via molecular dynamics simulations further improves predictions to outperform any other prediction method tested during the latest round of CASP. The resulting models have highly accurate global and local structures, including high accuracy at functionally important interface residues, and they are highly suitable as initial models for crystal structure determination via molecular replacement.

Ligand-Binding-Site Structure Refinement Using Molecular Dynamics with Restraints Derived from Predicted Binding Site Templates

Accurate modeling of ligand-binding-site structures plays a critical role in structure-based virtual screening. However, the structures of the ligand-binding site in most predicted protein models are generally of low quality and need refinements. In this work, we present a ligand-binding-site structure refinement protocol using molecular dynamics simulation with restraints derived from predicted binding site templates. Our benchmark validation shows great performance for 40 diverse sets of proteins from the Astex list. The ligand-binding sites on modeled protein structures are consistently refined using our method with an average Cα RMSD improvement of 0.90 Å. Comparison of ligand binding modes from ligand docking to initial unrefined and refined structures shows an average of 1.97 Å RMSD improvement in the refined structures. These results demonstrate a promising new method of structure refinement for protein ligand-binding-site structures.

Novel phosphorylation-dependent regulation in an unstructured protein

This work explores how phosphorylation of an unstructured protein region in inhibitor-2 (I2) regulates protein phosphatase-1 (PP1) enzyme activity using molecular dynamics (MD). Free I2 is largely unstructured; however, when bound to PP1, three segments adopt a stable structure. In particular, an I2 helix (i-helix) blocks the PP1 active site and inhibits phosphatase activity. I2 phosphorylation in the PP1-I2 complex activates phosphatase activity without I2 dissociation. The I2 Thr74 regulatory phosphorylation site is in an unstructured domain in PP1-I2. PP1-I2 MD demonstrated that I2 phosphorylation promotes early steps of PP1-I2 activation in explicit solvent models. Moreover, phosphorylation-dependent activation occurred in PP1-I2 complexes derived from I2 orthologs with diverse sequences from human, yeast, worm, and protozoa. This system allowed exploration of features of the 73-residue unstructured human I2 domain critical for phosphorylation-dependent activation. These studies revealed that components of I2 unstructured domain are strategically positioned for phosphorylation responsiveness including a transient α-helix. There was no evidence that electrostatic interactions of I2 phosphothreonine74 influenced PP1-I2 activation. Instead, phosphorylation altered the conformation of residues around Thr74. Phosphorylation uncurled the distance between I2 residues Glu71 to Tyr76 to promote PP1-I2 activation, whereas reduced distances reduced activation. This I2 residue Glu71 to Tyr76 distance distribution, independently from Thr74 phosphorylation, controls I2 i-helix displacement from the PP1 active site leading to PP1-I2 activation.

High-accuracy refinement using Rosetta in CASP13

Because proteins generally fold to their lowest free energy states, energy-guided refinement in principle should be able to systematically improve the quality of protein structure models generated using homologous structure or co-evolution derived information. However, because of the high dimensionality of the search space, there are far more ways to degrade the quality of a near native model than to improve it, and hence, refinement methods are very sensitive to energy function errors. In the 13th Critial Assessment of techniques for protein Structure Prediction (CASP13), we sought to carry out a thorough search for low energy states in the neighborhood of a starting model using restraints to avoid straying too far. The approach was reasonably successful in improving both regions largely incorrect in the starting models as well as core regions that started out closer to the correct structure. Models with GDT-HA over 70 were obtained for five targets and for one of those, an accuracy of 0.5 å backbone root-mean-square deviation (RMSD) was achieved. An important current challenge is to improve performance in refining oligomers and larger proteins, for which the search problem remains extremely difficult.

Applications of Molecular Dynamics Simulation in Structure Prediction of Peptides and Proteins

Compared with rapid accumulation of protein sequences from high-throughput DNA sequencing, obtaining experimental 3D structures of proteins is still much more difficult, making protein structure prediction (PSP) potentially very useful. Currently, a vast majority of PSP efforts are based on data mining of known sequences, structures and their relationships (informatics-based). However, if closely related template is not available, these methods are usually much less reliable than experiments. They may also be problematic in predicting the structures of naturally occurring or designed peptides. On the other hand, physics-based methods including molecular dynamics (MD) can utilize our understanding of detailed atomic interactions determining biomolecular structures. In this mini-review, we show that all-atom MD can predict structures of cyclic peptides and other peptide foldamers with accuracy similar to experiments. Then, some notable successes in reproducing experimental 3D structures of small proteins through MD simulations (some with replica-exchange) of the folding were summarized. We also describe advancements of MD-based refinement of structure models, and the integration of limited experimental or bioinformatics data into MD-based structure modeling.

Driven to near-experimental accuracy by refinement via molecular dynamics simulations

Protein model refinement has been an essential part of successful protein structure prediction. Molecular dynamics simulation-based refinement methods have shown consistent improvement of protein models. There had been progress in the extent of refinement for a few years since the idea of ensemble averaging of sampled conformations emerged. There was little progress in CASP12 because conformational sampling was not sufficiently diverse due to harmonic restraints. During CASP13, a new refinement method was tested that achieved significant improvements over CASP12. The new method intended to address previous bottlenecks in the refinement problem by introducing new features. Flat-bottom harmonic restraints replaced harmonic restraints, sampling was performed iteratively, and a new scoring function and selection criteria were used. The new protocol expanded conformational sampling at reduced computational costs. In addition to overall improvements, some models were refined significantly to near-experimental accuracy.

ReFOLD: a server for the refinement of 3D protein models guided by accurate quality estimates

ReFOLD is a novel hybrid refinement server with integrated high performance global and local Accuracy Self Estimates (ASEs). The server attempts to identify and to fix likely errors in user supplied 3D models of proteins via successive rounds of refinement. The server is unique in providing output for multiple alternative refined models in a way that allows users to quickly visualize the key residue locations, which are likely to have been improved. This is important, as global refinement of a full chain model may not always be possible, whereas local regions, or individual domains, can often be much improved. Thus, users may easily compare the specific regions of the alternative refined models in which they are most interested e.g. key interaction sites or domains. ReFOLD was used to generate hundreds of alternative refined models for the CASP12 experiment, boosting our group's performance in the main tertiary structure prediction category. Our successful refinement of initial server models combined with our built-in ASEs were instrumental to our second place ranking on Template Based Modeling (TBM) and Free Modeling (FM)/TBM targets. The ReFOLD server is freely available at: http://www.reading.ac.uk/bioinf/ReFOLD/.

Methods for the Refinement of Protein Structure 3D Models

The refinement of predicted 3D protein models is crucial in bringing them closer towards experimental accuracy for further computational studies. Refinement approaches can be divided into two main stages: The sampling and scoring stages. Sampling strategies, such as the popular Molecular Dynamics (MD)-based protocols, aim to generate improved 3D models. However, generating 3D models that are closer to the native structure than the initial model remains challenging, as structural deviations from the native basin can be encountered due to force-field inaccuracies. Therefore, different restraint strategies have been applied in order to avoid deviations away from the native structure. For example, the accurate prediction of local errors and/or contacts in the initial models can be used to guide restraints. MD-based protocols, using physics-based force fields and smart restraints, have made significant progress towards a more consistent refinement of 3D models. The scoring stage, including energy functions and Model Quality Assessment Programs (MQAPs) are also used to discriminate near-native conformations from non-native conformations. Nevertheless, there are often very small differences among generated 3D models in refinement pipelines, which makes model discrimination and selection problematic. For this reason, the identification of the most native-like conformations remains a major challenge.

Higher Accuracy Achieved in the Simulations of Protein Structure Refinement, Protein Folding, and Intrinsically Disordered Proteins Using Polarizable Force Fields

The accuracy of molecular mechanics force fields is of vital importance in biomolecular simulations. However, the admittedly more accurate polarizable force fields were recently reported to be less able to reproduce the experimental properties in comparison to additive force fields in some cases. Here, we perform long-time-scale molecular dynamics simulations to systematically evaluate the effect of explicit electronic polarization in polarizable force fields. The results show that the inclusion of electrostatic polarization effect in polarizable force fields can improve their accuracies in protein structure refinement and generate conformational ensembles more approximate to experiments for intrinsically disordered proteins. In contrast, it is difficult for polarizable force fields to approach the native structure, let alone to predict the native state when it is unknown a priori in the real protein structure predictions. We speculate that these effects might be attributed to the preference of protein-water interactions in polarizable force fields.

Experimental accuracy in protein structure refinement via molecular dynamics simulations

Refinement is the last step in protein structure prediction pipelines to convert approximate homology models to experimental accuracy. Protocols based on molecular dynamics (MD) simulations have shown promise, but current methods are limited to moderate levels of consistent refinement. To explore the energy landscape between homology models and native structures and analyze the challenges of MD-based refinement, eight test cases were studied via extensive simulations followed by Markov state modeling. In all cases, native states were found very close to the experimental structures and at the lowest free energies, but refinement was hindered by a rough energy landscape. Transitions from the homology model to the native states require the crossing of significant kinetic barriers on at least microsecond time scales. A significant energetic driving force toward the native state was lacking until its immediate vicinity, and there was significant sampling of off-pathway states competing for productive refinement. The role of recent force field improvements is discussed and transition paths are analyzed in detail to inform which key transitions have to be overcome to achieve successful refinement.

Modeling of Protein Structural Flexibility and Large-Scale Dynamics: Coarse-Grained Simulations and Elastic Network Models

Fluctuations of protein three-dimensional structures and large-scale conformational transitions are crucial for the biological function of proteins and their complexes. Experimental studies of such phenomena remain very challenging and therefore molecular modeling can be a good alternative or a valuable supporting tool for the investigation of large molecular systems and long-time events. In this minireview, we present two alternative approaches to the coarse-grained (CG) modeling of dynamic properties of protein systems. We discuss two CG representations of polypeptide chains used for Monte Carlo dynamics simulations of protein local dynamics and conformational transitions, and highly simplified structure-based elastic network models of protein flexibility. In contrast to classical all-atom molecular dynamics, the modeling strategies discussed here allow the quite accurate modeling of much larger systems and longer-time dynamic phenomena. We briefly describe the main features of these models and outline some of their applications, including modeling of near-native structure fluctuations, sampling of large regions of the protein conformational space, or possible support for the structure prediction of large proteins and their complexes.

Molecular Dynamics Simulation for All

The impact of molecular dynamics (MD) simulations in molecular biology and drug discovery has expanded dramatically in recent years. These simulations capture the behavior of proteins and other biomolecules in full atomic detail and at very fine temporal resolution. Major improvements in simulation speed, accuracy, and accessibility, together with the proliferation of experimental structural data, have increased the appeal of biomolecular simulation to experimentalists-a trend particularly noticeable in, although certainly not limited to, neuroscience. Simulations have proven valuable in deciphering functional mechanisms of proteins and other biomolecules, in uncovering the structural basis for disease, and in the design and optimization of small molecules, peptides, and proteins. Here we describe, in practical terms, the types of information MD simulations can provide and the ways in which they typically motivate further experimental work.

Structure refinement of membrane proteins via molecular dynamics simulations

A refinement protocol based on physics-based techniques established for water soluble proteins is tested for membrane protein structures. Initial structures were generated by homology modeling and sampled via molecular dynamics simulations in explicit lipid bilayer and aqueous solvent systems. Snapshots from the simulations were selected based on scoring with either knowledge-based or implicit membrane-based scoring functions and averaged to obtain refined models. The protocol resulted in consistent and significant refinement of the membrane protein structures similar to the performance of refinement methods for soluble proteins. Refinement success was similar between sampling in the presence of lipid bilayers and aqueous solvent but the presence of lipid bilayers may benefit the improvement of lipid-facing residues. Scoring with knowledge-based functions (DFIRE and RWplus) was found to be as good as scoring using implicit membrane-based scoring functions suggesting that differences in internal packing is more important than orientations relative to the membrane during the refinement of membrane protein homology models.

Protein homology model refinement by large-scale energy optimization

Proteins fold to their lowest free-energy structures, and hence the most straightforward way to increase the accuracy of a partially incorrect protein structure model is to search for the lowest-energy nearby structure. This direct approach has met with little success for two reasons: first, energy function inaccuracies can lead to false energy minima, resulting in model degradation rather than improvement; and second, even with an accurate energy function, the search problem is formidable because the energy only drops considerably in the immediate vicinity of the global minimum, and there are a very large number of degrees of freedom. Here we describe a large-scale energy optimization-based refinement method that incorporates advances in both search and energy function accuracy that can substantially improve the accuracy of low-resolution homology models. The method refined low-resolution homology models into correct folds for 50 of 84 diverse protein families and generated improved models in recent blind structure prediction experiments. Analyses of the basis for these improvements reveal contributions from both the improvements in conformational sampling techniques and the energy function.

PREFMD: a web server for protein structure refinement via molecular dynamics simulations

Summary

Refinement of protein structure models is a long-standing problem in structural bioinformatics. Molecular dynamics-based methods have emerged as an avenue to achieve consistent refinement. The PREFMD web server implements an optimized protocol based on the method successfully tested in CASP11. Validation with recent CASP refinement targets shows consistent and more significant improvement in global structure accuracy over other state-of-the-art servers.

Availability and implementation

PREFMD is freely available as a web server at http://feiglab.org/prefmd. Scripts for running PREFMD as a stand-alone package are available at https://github.com/feiglab/prefmd.git.

Contact

feig@msu.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

What makes it difficult to refine protein models further via molecular dynamics simulations?

Protein structure refinement remains a challenging yet important problem as it has the potential to bring already accurate template-based models to near-native resolution. Refinement based on molecular dynamics simulations has been a highly promising approach and the performance of MD-based refinement in the Feig group during CASP12 is described here. During CASP12, sampling was extended well into the microsecond scale, an improved force field was applied, and new protocol variations were tested. Progress over previous rounds of CASP was found to be limited which is analyzed in terms of the quality of the initial models and dependency on the amount of sampling and refinement protocol variations. As current MD-based refinement protocols appear to be reaching a plateau, detailed analysis is presented to provide new insight into the major challenges towards more extensive structure refinement, focusing in particular on sampling with and without restraints.

Principal component analysis in protein tertiary structure prediction

We discuss applicability of principal component analysis (PCA) for protein tertiary structure prediction from amino acid sequence. The algorithm presented in this paper belongs to the category of protein refinement models and involves establishing a low-dimensional space where the sampling (and optimization) is carried out via particle swarm optimizer (PSO). The reduced space is found via PCA performed for a set of low-energy protein models previously found using different optimization techniques. A high frequency term is added into this expansion by projecting the best decoy into the PCA basis set and calculating the residual model. This term is aimed at providing high frequency details in the energy optimization. The goal of this research is to analyze how the dimensionality reduction affects the prediction capability of the PSO procedure. For that purpose, different proteins from the Critical Assessment of Techniques for Protein Structure Prediction experiments were modeled. In all the cases, both the energy of the best decoy and the distance to the native structure have decreased. Our analysis also shows how the predicted backbone structure of native conformation and of alternative low energy states varies with respect to the PCA dimensionality. Generally speaking, the reconstruction can be successfully achieved with 10 principal components and the high frequency term. We also provide a computational analysis of protein energy landscape for the inverse problem of reconstructing structure from the reduced number of principal components, showing that the dimensionality reduction alleviates the ill-posed character of this high-dimensional energy optimization problem. The procedure explained in this paper is very fast and allows testing different PCA expansions. Our results show that PSO improves the energy of the best decoy used in the PCA when the adequate number of PCA terms is considered.

Protein structure modeling and refinement by global optimization in CASP12

For protein structure modeling in the CASP12 experiment, we have developed a new protocol based on our previous CASP11 approach. The global optimization method of conformational space annealing (CSA) was applied to 3 stages of modeling: multiple sequence-structure alignment, three-dimensional (3D) chain building, and side-chain re-modeling. For better template selection and model selection, we updated our model quality assessment (QA) method with the newly developed SVMQA (support vector machine for quality assessment). For 3D chain building, we updated our energy function by including restraints generated from predicted residue-residue contacts. New energy terms for the predicted secondary structure and predicted solvent accessible surface area were also introduced. For difficult targets, we proposed a new method, LEEab, where the template term played a less significant role than it did in LEE, complemented by increased contributions from other terms such as the predicted contact term. For TBM (template-based modeling) targets, LEE performed better than LEEab, but for FM targets, LEEab was better. For model refinement, we modified our CASP11 molecular dynamics (MD) based protocol by using explicit solvents and tuning down restraint weights. Refinement results from MD simulations that used a new augmented statistical energy term in the force field were quite promising. Finally, when using inaccurate information (such as the predicted contacts), it was important to use the Lorentzian function for which the maximal penalty arising from wrong information is always bounded.

Assessment of the model refinement category in CASP12

We here report on the assessment of the model refinement predictions submitted to the 12th Experiment on the Critical Assessment of Protein Structure Prediction (CASP12). This is the fifth refinement experiment since CASP8 (2008) and, as with the previous experiments, the predictors were invited to refine selected server models received in the regular (nonrefinement) stage of the CASP experiment. We assessed the submitted models using a combination of standard CASP measures. The coefficients for the linear combination of Z-scores (the CASP12 score) have been obtained by a machine learning algorithm trained on the results of visual inspection. We identified eight groups that improve both the backbone conformation and the side chain positioning for the majority of targets. Albeit the top methods adopted distinctively different approaches, their overall performance was almost indistinguishable, with each of them excelling in different scores or target subsets. What is more, there were a few novel approaches that, while doing worse than average in most cases, provided the best refinements for a few targets, showing significant latitude for further innovation in the field.

Structural insights into DNA cleavage activation of CRISPR-Cas9 system

CRISPR-Cas9 technology has been widely used for genome engineering. Its RNA-guided endonuclease Cas9 binds specifically to target DNA and then cleaves the two DNA strands with HNH and RuvC nuclease domains. However, structural information regarding the DNA cleavage-activating state of two nuclease domains remains sparse. Here, we report a 5.2 Å cryo-EM structure of Cas9 in complex with sgRNA and target DNA. This structure reveals a conformational state of Cas9 in which the HNH domain is closest to the DNA cleavage site. Compared with two known HNH states, our structure shows that the HNH active site moves toward the cleavage site by about 25 and 13 Å, respectively. In combination with EM-based molecular dynamics simulations, we show that residues of the nuclease domains in our structure could form cleavage-compatible conformations with the target DNA. Together, these results strongly suggest that our cryo-EM structure resembles a DNA cleavage-activating architecture of Cas9.

Sampling and energy evaluation challenges in ligand binding protein design

The steroid hormone 17α-hydroxylprogesterone (17-OHP) is a biomarker for congenital adrenal hyperplasia and hence there is considerable interest in development of sensors for this compound. We used computational protein design to generate protein models with binding sites for 17-OHP containing an extended, nonpolar, shape-complementary binding pocket for the four-ring core of the compound, and hydrogen bonding residues at the base of the pocket to interact with carbonyl and hydroxyl groups at the more polar end of the ligand. Eight of 16 designed proteins experimentally tested bind 17-OHP with micromolar affinity. A co-crystal structure of one of the designs revealed that 17-OHP is rotated 180° around a pseudo-two-fold axis in the compound and displays multiple binding modes within the pocket, while still interacting with all of the designed residues in the engineered site. Subsequent rounds of mutagenesis and binding selection improved the ligand affinity to nanomolar range, while appearing to constrain the ligand to a single bound conformation that maintains the same "flipped" orientation relative to the original design. We trace the discrepancy in the design calculations to two sources: first, a failure to model subtle backbone changes which alter the distribution of sidechain rotameric states and second, an underestimation of the energetic cost of desolvating the carbonyl and hydroxyl groups of the ligand. The difference between design model and crystal structure thus arises from both sampling limitations and energy function inaccuracies that are exacerbated by the near two-fold symmetry of the molecule.

Simultaneous refinement of inaccurate local regions and overall structure in the CASP12 protein model refinement experiment

Advances in protein model refinement techniques are required as diverse sources of protein structure information are available from low-resolution experiments or informatics-based computations such as cryo-EM, NMR, homology models, or predicted residue contacts. Given semi-reliable or incomplete structural information, structure quality of a protein model has to be improved by ab initio methods such as energy-based simulation. In this study, we describe a new automatic refinement server method designed to improve locally inaccurate regions and overall structure simultaneously. Locally inaccurate regions may occur in protein structures due to non-convergent or missing information in template structures used in homology modeling or due to intrinsic structural flexibilities not resolved by experimental techniques. However, such variable or dynamic regions often play important functional roles by participating in interactions with other biomolecules or in transitions between different functional states. The new refinement method introduced here utilizes diverse types of geometric operators which drive both local and global changes, and the effect of structure changes and relaxations are accumulated. This resulted in consistent refinement of both local and global structural features. Performance of this method in CASP12 is discussed.

Protein structure prediction using Rosetta in CASP12

We describe several notable aspects of our structure predictions using Rosetta in CASP12 in the free modeling (FM) and refinement (TR) categories. First, we had previously generated (and published) models for most large protein families lacking experimentally determined structures using Rosetta guided by co-evolution based contact predictions, and for several targets these models proved better starting points for comparative modeling than any known crystal structure-our model database thus starts to fulfill one of the goals of the original protein structure initiative. Second, while our "human" group simply submitted ROBETTA models for most targets, for six targets expert intervention improved predictions considerably; the largest improvement was for T0886 where we correctly parsed two discontinuous domains guided by predicted contact maps to accurately identify a structural homolog of the same fold. Third, Rosetta all atom refinement followed by MD simulations led to consistent but small improvements when starting models were close to the native structure, and larger but less consistent improvements when starting models were further away.

A Simple and Efficient Protein Structure Refinement Method

Improving the quality of a given protein structure can serve as the ultimate solution for accurate protein structure prediction, and seeking such a method is currently a challenge in computational structural biology. In order to promote and encourage much needed such efforts, CASP (Critical Assessment of Structure Prediction) has been providing an ideal computational experimental platform, where it was reported only recently (since CASP10) that systematic protein structure refinement is possible by carrying out extensive (approximately millisecond) MD simulations with proper restraints generated from the given structure. Using an explicit solvent model and much reduced positional and distance restraints than previously exercised, we propose a refinement protocol that combines a series of short (5 ns) MD simulations with energy minimization procedures. Testing and benchmarking on 54 CASP8-10 refinement targets and 34 CASP11 refinement targets shows quite promising results. Using only a small fraction of MD simulation steps (nanosecond versus millisecond), systematic protein structure refinement was demonstrated in this work, indicating that refinement of a given model can be achieved using a few hours of desktop computing.

All atom NMDA receptor transmembrane domain model development and simulations in lipid bilayers and water

N-methyl-d-aspartate receptors (NMDARs) are members of the ionotropic glutamate receptor family that mediate excitatory synaptic transmission in the central nervous system. The channels of NMDARs are permeable to Ca²⁺ but blocked by Mg²⁺, distinctive properties that underlie essential brain processes such as induction of synaptic plasticity. However, due to limited structural information about the NMDAR transmembrane ion channel forming domain, the mechanism of divalent cation permeation and block is understood poorly. In this paper we developed an atomistic model of the transmembrane domain (TMD) of NMDARs composed of GluN1 and GluN2A subunits (GluN1/2A receptors). The model was generated using (a) a homology model based on the structure of the NaK channel and a partially resolved structure of an AMPA receptor (AMPAR), and (b) a partially resolved X-ray structure of GluN1/2B NMDARs. Refinement and extensive Molecular Dynamics (MD) simulations of the NMDAR TMD model were performed in explicit lipid bilayer membrane and water. Targeted MD with simulated annealing was introduced to promote structure refinement. Putative positions of the Mg²⁺ and Ca²⁺ ions in the ion channel divalent cation binding site are proposed. Differences in the structural and dynamic behavior of the channel protein in the presence of Mg²⁺ or Ca²⁺ are analyzed. NMDAR protein conformational flexibility was similar with no ion bound to the divalent cation binding site and with Ca²⁺ bound, whereas Mg²⁺ binding reduced protein fluctuations. While bound at the binding site both ions retained their preferred ligand coordination numbers: 6 for Mg²⁺, and 7–8 for Ca²⁺. Four asparagine side chain oxygens, a back-bone oxygen, and a water molecule participated in binding a Mg²⁺ ion. The Ca²⁺ ion first coordination shell ligands typically included four to five side-chain oxygen atoms of the binding site asparagine residues, two water molecules and zero to two backbone oxygens of the GluN2B subunits. These results demonstrate the importance of high-resolution channel structures for elucidation of mechanisms of NMDAR permeation and block.

Computational protein structure refinement: Almost there, yet still so far to go

Protein structures are essential in modern biology yet experimental methods are far from being able to catch up with the rapid increase in available genomic data. Computational protein structure prediction methods aim to fill the gap while the role of protein structure refinement is to take approximate initial template-based models and bring them closer to the true native structure. Current methods for computational structure refinement rely on molecular dynamics simulations, related sampling methods, or iterative structure optimization protocols. The best methods are able to achieve moderate degrees of refinement but consistent refinement that can reach near-experimental accuracy remains elusive. Key issues revolve around the accuracy of the energy function, the inability to reliably rank multiple models, and the use of restraints that keep sampling close to the native state but also limit the degree of possible refinement. A different aspect is the question of what exactly the target of high-resolution refinement should be as experimental structures are affected by experimental conditions and different biological questions require varying levels of accuracy. While improvement of the global protein structure is a difficult problem, high-resolution refinement methods that improves local structural quality such as favorable stereochemistry and the avoidance of atomic clashes are much more successful.

Accelerating physical simulations of proteins by leveraging external knowledge

It is challenging to compute structure-function relationships of proteins using molecular physics. The problem arises from the exponential scaling of the computational searching and sampling of large conformational spaces. This scaling challenge is not met by today's methods, such as Monte Carlo, simulated annealing, genetic algorithms, or molecular dynamics (MD) or its variants such as replica exchange. Such methods of searching for optimal states on complex probabalistic landscapes are referred to more broadly as Explore-and-Exploit (EE), including in contexts such as computational learning, games, industrial planning and modeling military strategies. Here we describe a Bayesian method, called MELD, that 'melds' together explore-and-exploit approaches with externally added information that can be vague, combinatoric, noisy, intuitive, heuristic, or from experimental data. MELD is shown to accelerate physical MD simulations when using experimental data to determine protein structures; for predicting protein structures by using heuristic directives; and when predicting binding affinities of proteins from limited information about the binding site. Such Guided Explore-and-Exploit approaches might also be useful beyond proteins and beyond molecular science.

Computational refinement of spectroscopic FRET measurements

This article supplies raw data related to a research article entitled “Joint refinement of FRET measurements using spectroscopic and computational tools” (Kyrychenko et al., 2017) [1], in which we demonstrate the use of molecular dynamics simulations to estimate FRET orientational factors in a benchmark donor-linker-acceptor system of enhanced cyan (ECFP) and enhanced yellow (EYFP) fluorescent proteins. This can improve the recalculation of donor-acceptor distance information from single-molecule FRET measurements.

Princeton_TIGRESS 2.0: High refinement consistency and net gains through support vector machines and molecular dynamics in double-blind predictions during the CASP11 experiment

Protein structure refinement is the challenging problem of operating on any protein structure prediction to improve its accuracy with respect to the native structure in a blind fashion. Although many approaches have been developed and tested during the last four CASP experiments, a majority of the methods continue to degrade models rather than improve them. Princeton_TIGRESS (Khoury et al., Proteins 2014;82:794-814) was developed previously and utilizes separate sampling and selection stages involving Monte Carlo and molecular dynamics simulations and classification using an SVM predictor. The initial implementation was shown to consistently refine protein structures 76% of the time in our own internal benchmarking on CASP 7-10 targets. In this work, we improved the sampling and selection stages and tested the method in blind predictions during CASP11. We added a decomposition of physics-based and hybrid energy functions, as well as a coordinate-free representation of the protein structure through distance-binning Cα-Cα distances to capture fine-grained movements. We performed parameter estimation to optimize the adjustable SVM parameters to maximize precision while balancing sensitivity and specificity across all cross-validated data sets, finding enrichment in our ability to select models from the populations of similar decoys generated for targets in CASPs 7-10. The MD stage was enhanced such that larger structures could be further refined. Among refinement methods that are currently implemented as web-servers, Princeton_TIGRESS 2.0 demonstrated the most consistent and most substantial net refinement in blind predictions during CASP11. The enhanced refinement protocol Princeton_TIGRESS 2.0 is freely available as a web server at http://atlas.engr.tamu.edu/refinement/. Proteins 2017; 85:1078-1098. © 2017 Wiley Periodicals, Inc.

Joint refinement of FRET measurements using spectroscopic and computational tools

The variability of the orientation factor is a long-standing challenge in converting FRET efficiency measurements into donor-acceptor distances. We propose the use of molecular dynamics (MD) simulations to characterize orientation distributions and thus improve the accuracy of distance measurements. Here, we test this approach by comparing experimental and simulated FRET efficiencies for a model donor-acceptor pair of enhanced cyan and enhanced yellow FPs connected by a flexible linker. Several spectroscopic techniques were used to characterize FRET in solution. In addition, a series of atomistic MD simulations of a total length of 1.5 μs were carried out to calculate the distances and the orientation factor in the FRET-pair. The resulting MD-based and experimentally measured FRET efficiency histograms coincided with each other, allowing for direct comparison of distance distributions. Despite the fact that the calculated average orientation factor was close to 2/3, the application of the average κ² to the entire histogram of FRET efficiencies resulted in a substantial artificial broadening of the calculated distribution of apparent donor-acceptor distances. By combining single pair-FRET measurements with computational tools, we demonstrate that accounting for the donor and acceptor orientation heterogeneity is critical for accurate representation of the donor-acceptor distance distribution from FRET measurements.

Blind protein structure prediction using accelerated free-energy simulations

We report a key proof of principle of a new acceleration method [Modeling Employing Limited Data (MELD)] for predicting protein structures by molecular dynamics simulation. It shows that such Boltzmann-satisfying techniques are now sufficiently fast and accurate to predict native protein structures in a limited test within the Critical Assessment of Structure Prediction (CASP) community-wide blind competition.

Pharmacophore Models Derived from Molecular Dynamics Simulations of Protein-Ligand Complexes: A Case Study

A single, merged pharmacophore hypothesis is derived combining 2000 pharmacophore models obtained during a 20 ns molecular dynamics simulation of a protein-ligand complex with one pharmacophore model derived from the initial PDB structure. This merged pharmacophore model contains all features that are present during the simulation and statistical information about the dynamics of the pharmacophore features. Based on the dynamics of the pharmacophore features we derive two distinctive feature patterns resulting in two different pharmacophore models for the analyzed system – the first model consists of features that are obtained from the PDB structure and the second uses two features that can only be derived from the molecular dynamics simulation. Both models can distinguish between active and decoy molecules in virtual screening. Our approach represents an objective way to add/remove features in pharmacophore models and can be of interest for the investigation of any naturally occurring system that relies on ligand-receptor interactions for its biological activity.

Local Protein Structure Refinement via Molecular Dynamics Simulations with locPREFMD

A method for the local refinement of protein structures that targets improvements in local stereochemistry while preserving the overall fold is presented. The method uses force field-based minimization and sampling via molecular dynamics simulations with a modified force field to bring bonds, angles, and torsion angles into an acceptable range for high-resolution protein structures. The method is implemented in the locPREFMD web server and was tested on computational models submitted to CASP11. Using MolProbity scores as the main assessment criterion, the locPREFMD method significantly improves the stereochemical quality of given input models close to the quality expected for experimental structures while maintaining the Cα coordinates of the initial model.

Comparing pharmacophore models derived from crystal structures and from molecular dynamics simulations

ABSTRACT

Pharmacophore modeling is a widely used technique in computer-aided drug discovery. Structure-based pharmacophore models of a ligand in complex with a protein have proven to be useful for supporting in silico hit discovery, hit to lead expansion, and lead optimization. As a structure-based approach it depends on the correct interpretation of ligand-protein interactions. There are legitimate concerns about the fidelity of the bound ligand and about non-physiological contacts with parts of the crystal and the solvent effects that influence the protein structure. A possible way to refine the structure of a protein-ligand system is to use the final structure of a given MD simulation. In this study we compare pharmacophore models built using the initial protein-ligand structure obtained from the protein data bank (PDB) with pharmacophore models built with the final structure of a molecular dynamics simulation. We show that the pharmacophore models differ in feature number and feature type and that the pharmacophore models built from the last structure of a MD simulation shows in some cases better ability to distinguish between active and decoy ligand structures.

GRAPHICAL ABSTRACT