Forging tools for refining predicted protein structures

DynamicBind: predicting ligand-specific protein-ligand complex structure with a deep equivariant generative model

While significant advances have been made in predicting static protein structures, the inherent dynamics of proteins, modulated by ligands, are crucial for understanding protein function and facilitating drug discovery. Traditional docking methods, frequently used in studying protein-ligand interactions, typically treat proteins as rigid. While molecular dynamics simulations can propose appropriate protein conformations, they're computationally demanding due to rare transitions between biologically relevant equilibrium states. In this study, we present DynamicBind, a deep learning method that employs equivariant geometric diffusion networks to construct a smooth energy landscape, promoting efficient transitions between different equilibrium states. DynamicBind accurately recovers ligand-specific conformations from unbound protein structures without the need for holo-structures or extensive sampling. Remarkably, it demonstrates state-of-the-art performance in docking and virtual screening benchmarks. Our experiments reveal that DynamicBind can accommodate a wide range of large protein conformational changes and identify cryptic pockets in unseen protein targets. As a result, DynamicBind shows potential in accelerating the development of small molecules for previously undruggable targets and expanding the horizons of computational drug discovery.

Molecular frustration: a hypothesis for regulation of viral infections

The recent revolution in imaging techniques and results from RNA footprinting in situ reveal how the bacteriophage MS2 genome regulates both particle assembly and genome release. We have proposed a model in which multiple packaging signal (PS) RNA-coat protein (CP) contacts orchestrate different stages of a viral life cycle. Programmed formation and release of specific PS contacts with CP regulates viral particle assembly and genome uncoating during cell entry. We hypothesize that molecular frustration, a concept introduced to understand protein folding, can be used to better rationalize how PSs function in both particle assembly and genome release. More broadly this concept may explain the directionality of viral life cycles, for example, the roles of host cofactors in HIV infection. We propose that this is a universal principle in virology that explains mechanisms of host-virus interaction and suggests diverse therapeutic interventions.

Computationally exploring the mechanism of bacteriophage T7 gp4 helicase translocating along ssDNA

Bacteriophage T7 gp4 helicase has served as a model system for understanding mechanisms of hexameric replicative helicase translocation. The mechanistic basis of how nucleoside 5'-triphosphate hydrolysis and translocation of gp4 helicase are coupled is not fully resolved. Here, we used a thermodynamically benchmarked coarse-grained protein force field, Associative memory, Water mediated, Structure and Energy Model (AWSEM), with the single-stranded DNA (ssDNA) force field 3SPN.2C to investigate gp4 translocation. We found that the adenosine 5'-triphosphate (ATP) at the subunit interface stabilizes the subunit-subunit interaction and inhibits subunit translocation. Hydrolysis of ATP to adenosine 5'-diphosphate enables the translocation of one subunit, and new ATP binding at the new subunit interface finalizes the subunit translocation. The LoopD2 and the N-terminal primase domain provide transient protein-protein and protein-DNA interactions that facilitate the large-scale subunit movement. The simulations of gp4 helicase both validate our coarse-grained protein-ssDNA force field and elucidate the molecular basis of replicative helicase translocation.

Biomolecular Modeling and Simulation: A Prospering Multidisciplinary Field

We reassess progress in the field of biomolecular modeling and simulation, following up on our perspective published in 2011. By reviewing metrics for the field's productivity and providing examples of success, we underscore the productive phase of the field, whose short-term expectations were overestimated and long-term effects underestimated. Such successes include prediction of structures and mechanisms; generation of new insights into biomolecular activity; and thriving collaborations between modeling and experimentation, including experiments driven by modeling. We also discuss the impact of field exercises and web games on the field's progress. Overall, we note tremendous success by the biomolecular modeling community in utilization of computer power; improvement in force fields; and development and application of new algorithms, notably machine learning and artificial intelligence. The combined advances are enhancing the accuracy andscope of modeling and simulation, establishing an exemplary discipline where experiment and theory or simulations are full partners.

Protein Structure Refinement Guided by Atomic Packing Frustration Analysis

Recent advances in machine learning, bioinformatics, and the understanding of the folding problem have enabled efficient predictions of protein structures with moderate accuracy, even for targets where there is little information from templates. All-atom molecular dynamics simulations provide a route to refine such predicted structures, but unguided atomistic simulations, even when lengthy in time, often fail to eliminate incorrect structural features that would prevent the structure from becoming more energetically favorable owing to the necessity of making large scale motions and to overcoming energy barriers for side chain repacking. In this study, we show that localizing packing frustration at atomic resolution by examining the statistics of the energetic changes that occur when the local environment of a site is changed allows one to identify the most likely locations of incorrect contacts. The global statistics of atomic resolution frustration in structures that have been predicted using various algorithms provide strong indicators of structural quality when tested over a database of 20 targets from previous CASP experiments. Residues that are more correctly located turn out to be more minimally frustrated than more poorly positioned sites. These observations provide a diagnosis of both global and local quality of predicted structures and thus can be used as guidance in all-atom refinement simulations of the 20 targets. Refinement simulations guided by atomic packing frustration turn out to be quite efficient and significantly improve the quality of the structures.

Surveying biomolecular frustration at atomic resolution

To function, biomolecules require sufficient specificity of interaction as well as stability to live in the cell while still being able to move. Thermodynamic stability of only a limited number of specific structures is important so as to prevent promiscuous interactions. The individual interactions in proteins, therefore, have evolved collectively to give funneled minimally frustrated landscapes but some strategic parts of biomolecular sequences located at specific sites in the structure have been selected to be frustrated in order to allow both motion and interaction with partners. We describe a framework efficiently to quantify and localize biomolecular frustration at atomic resolution by examining the statistics of the energy changes that occur when the local environment of a site is changed. The location of patches of highly frustrated interactions correlates with key biological locations needed for physiological function. At atomic resolution, it becomes possible to extend frustration analysis to protein-ligand complexes. At this resolution one sees that drug specificity is correlated with there being a minimally frustrated binding pocket leading to a funneled binding landscape. Atomistic frustration analysis provides a route for screening for more specific compounds for drug discovery.

The analysis of biomolecular frustration yielded insights into several aspects of protein behavior. Here the authors describe a framework to efficiently quantify and localize biomolecular frustration within proteins at atomic resolution, and observe that drug specificity is correlated with a minimally frustrated binding pocket leading to a funneled binding landscape.

Molecular-replacement phasing using predicted protein structures from AWSEM-Suite

The phase problem in X-ray crystallography arises from the fact that only the intensities, and not the phases, of the diffracting electromagnetic waves are measured directly. Molecular replacement can often estimate the relative phases of reflections starting with those derived from a template structure, which is usually a previously solved structure of a similar protein. The key factor in the success of molecular replacement is finding a good template structure. When no good solved template exists, predicted structures based partially on templates can sometimes be used to generate models for molecular replacement, thereby extending the lower bound of structural and sequence similarity required for successful structure determination. Here, the effectiveness is examined of structures predicted by a state-of-the-art prediction algorithm, the Associative memory, Water-mediated, Structure and Energy Model Suite (AWSEM-Suite), which has been shown to perform well in predicting protein structures in CASP13 when there is no significant sequence similarity to a solved protein or only very low sequence similarity to known templates. The performance of AWSEM-Suite structures in molecular replacement is discussed and the results show that AWSEM-Suite performs well in providing useful phase information, often performing better than I-TASSER-MR and the previous algorithm AWSEM-Template.

Unfolding cytochromes c-b562 and Rd apo b562

We have analyzed the early stages of unfolding of cytochromes c-b₅₆₂ (PDB ID: 2BC5) and Rd apo b₅₆₂ (PDB ID: 1YYJ). Our geometrical approach proceeds from an analysis of the crystal structure reported for each protein. We quantify, residue-by-residue and region-by-region, the spatial and angular changes in the structure as the protein denatures, and quantify differences that result from the seven residues that differ in the two proteins. Using two independent analyses, one based on spatial metrics and the second on angular metrics, we establish the order of unfolding of the five helices in cyt c-b₅₆₂ and the four helices in the apo protein. For the two helices nearest the N-terminal end of both proteins, the ones in the apo protein unfold first. For the two helices nearest the C-terminal end, the interior helix of the apo protein unfolds first, whereas the terminal helix of the holo protein unfolds first. Excluded-volume effects (repulsive interactions) are minimized in turning regions; the overall range in Δ values is Δ = 36.3 Å³ for cyt c-b₅₆₂ and Δ = 36.6 Å³ for the apo protein, whereas the span for all 20 amino acids is Δ = 167.7 Å³. As our work indicates that the interior helix of cytochrome c-b₅₆₂ is the first to fold, we suggest that this helix protects the heme from misligation, consistent with ultrafast folding over a minimally frustrated funneled landscape.

Protein Structure Prediction in CASP13 Using AWSEM-Suite

Recently several techniques have emerged that significantly enhance the quality of predictions of protein tertiary structures. In this study, we describe the performance of AWSEM-Suite, an algorithm that incorporates template-based modeling and coevolutionary restraints with a realistic coarse-grained force field, AWSEM. With its roots in neural networks, AWSEM contains both physical and bioinformatical energies that have been optimized using energy landscape theory. AWSEM-Suite participated in CASP13 as a server predictor and generated reliable predictions for most targets. AWSEM-Suite ranked eighth in both the free-modeling category and the hard-to-model category and in one case provided the best submitted prediction. Here we critically discuss the prediction performance of AWSEM-Suite using several examples from different categories in CASP13. Structure prediction tests on these selected targets, two of them being hard-to-model targets, show that AWSEM-Suite can achieve high-resolution structure prediction after incorporating both template guidances and coevolutionary restraints even when homology is weak. For targets with reliable templates (template-easy category), introducing coevolutionary restraints sometimes damages the overall quality of the predictions. Free energy profile analyses demonstrate, however, that the incorporations of both of these evolutionarily informed terms effectively increase the funneling of the landscape toward native-like structures while still allowing sufficient flexibility to correct for discrepancies between the correct target structure and the provided guidance. In contrast to other predictors that are exclusively oriented toward structure prediction, the connection of AWSEM-Suite to a statistical mechanical basis and affiliated molecular dynamics and importance sampling simulations makes it suitable for functional explorations.

Funneled angle landscapes for helical proteins

We use crystallographic data for four helical iron proteins (cytochrome c-b₅₆₂, cytochrome c', sperm whale myoglobin, human cytoglobin) to calculate radial and angular signatures as each unfolds from the native state stepwise though four unfolded states. From these data we construct an angle phase diagram to display the evolution of each protein from its native state; and, in turn, the phase diagram is used to construct a funneled angle landscape for comparison with the topography of its folding energy landscape. We quantify the departure of individual helical and turning regions from the areal, angular profile of corresponding regions of the native state. This procedure allows us to identify the similarities and differences among individual helical and turning regions in the early stages of unfolding of the four helical heme proteins.

Computational reconstruction of atomistic protein structures from coarse-grained models

Three-dimensional protein structures, whether determined experimentally or theoretically, are often too low resolution. In this mini-review, we outline the computational methods for protein structure reconstruction from incomplete coarse-grained to all atomistic models. Typical reconstruction schemes can be divided into four major steps. Usually, the first step is reconstruction of the protein backbone chain starting from the C-alpha trace. This is followed by side-chains rebuilding based on protein backbone geometry. Subsequently, hydrogen atoms can be reconstructed. Finally, the resulting all-atom models may require structure optimization. Many methods are available to perform each of these tasks. We discuss the available tools and their potential applications in integrative modeling pipelines that can transfer coarse-grained information from computational predictions, or experiment, to all atomistic structures.

Disorder Mediated Oligomerization of DISC1 Proteins Revealed by Coarse-Grained Molecular Dynamics Simulations

Disrupted-in-schizophrenia-1 (DISC1) is a scaffold protein of significant importance for neuro-development and a prominent candidate protein in the etiology of mental disorders. In this work, we investigate the role of conformational heterogeneity and local structural disorder in the oligomerization pathway of the full-length DISC1 and of two truncation variants. Through extensive coarse-grained molecular dynamics simulations with a predictive energy landscape-based model, we shed light on the interplay of local and global disorder which lead to different oligomerization pathways. We found that both global conformational heterogeneity and local structural disorder play an important role in shaping distinct oligomerization trends of DISC1. This study also sheds light on the differences in oligomerization pathways of the full-length protein compared to the truncated variants produced by a chromosomal translocation associated with schizophrenia. We report that oligomerization of full-length DISC1 sequence works in a nonadditive manner with respect to truncated fragments that do not mirror the conformational landscape or binding affinities of the full-length unit.

Stereochemistry of residues in turning regions of helical proteins

We have developed a geometrical approach to quantify differences in the stereochemistry of α-helical and turning regions in four iron proteins. Two spatial signatures are used to analyze residue coordinate data for each protein; and a third is employed to analyze amino-acid molecular volume data. The residue-by-residue analysis of the results, taken together with the finding that two major factors stabilize an α-helix (minimization of side-chain steric interference and intrachain H-bonding), lead to the conclusion that certain residues are preferentially selected for α-helix formation. In the sequential, de novo synthesis of a turning region, residues are preferentially selected such that the overall molecular volume profile (representing purely repulsive, excluded-volume effects) spans a small range Δ of values (Δ = 39.1 Å³) relative to the total range that could be spanned (Δ = 167.7 Å³). It follows that excluded-volume effects are of enormous importance for residues in helical regions as well as those in adjacent turning regions. Once steric effects are taken into account, down-range attractive interactions between residues come into play in the formation of α-helical regions. The geometry of α-helices can be accommodated by conformational changes in less-structured turning regions of a polypeptide, thereby producing a globally optimized (native) protein structure.