Community-wide assessment of protein-interface modeling suggests improvements to design methodology

A suite of designed protein cages using machine learning and protein fragment-based protocols

Designed protein cages and related materials provide unique opportunities for applications in biotechnology and medicine, but their creation remains challenging. Here, we apply computational approaches to design a suite of tetrahedrally symmetric, self-assembling protein cages. For the generation of docked conformations, we emphasize a protein fragment-based approach, while for sequence design of the de novo interface, a comparison of knowledge-based and machine learning protocols highlights the power and increased experimental success achieved using ProteinMPNN. An analysis of design outcomes provides insights for improving interface design protocols, including prioritizing fragment-based motifs, balancing interface hydrophobicity and polarity, and identifying preferred polar contact patterns. In all, we report five structures for seven protein cages, along with two structures of intermediate assemblies, with the highest resolution reaching 2.0 Å using cryo-EM. This set of designed cages adds substantially to the body of available protein nanoparticles, and to methodologies for their creation.

Impact of AlphaFold on structure prediction of protein complexes: The CASP15-CAPRI experiment

We present the results for CAPRI Round 54, the 5th joint CASP-CAPRI protein assembly prediction challenge. The Round offered 37 targets, including 14 homodimers, 3 homo-trimers, 13 heterodimers including 3 antibody-antigen complexes, and 7 large assemblies. On average ~70 CASP and CAPRI predictor groups, including more than 20 automatics servers, submitted models for each target. A total of 21 941 models submitted by these groups and by 15 CAPRI scorer groups were evaluated using the CAPRI model quality measures and the DockQ score consolidating these measures. The prediction performance was quantified by a weighted score based on the number of models of acceptable quality or higher submitted by each group among their five best models. Results show substantial progress achieved across a significant fraction of the 60+ participating groups. High-quality models were produced for about 40% of the targets compared to 8% two years earlier. This remarkable improvement is due to the wide use of the AlphaFold2 and AlphaFold2-Multimer software and the confidence metrics they provide. Notably, expanded sampling of candidate solutions by manipulating these deep learning inference engines, enriching multiple sequence alignments, or integration of advanced modeling tools, enabled top performing groups to exceed the performance of a standard AlphaFold2-Multimer version used as a yard stick. This notwithstanding, performance remained poor for complexes with antibodies and nanobodies, where evolutionary relationships between the binding partners are lacking, and for complexes featuring conformational flexibility, clearly indicating that the prediction of protein complexes remains a challenging problem.

A Suite of Designed Protein Cages Using Machine Learning Algorithms and Protein Fragment-Based Protocols

Designed protein cages and related materials provide unique opportunities for applications in biotechnology and medicine, while methods for their creation remain challenging and unpredictable. In the present study, we apply new computational approaches to design a suite of new tetrahedrally symmetric, self-assembling protein cages. For the generation of docked poses, we emphasize a protein fragment-based approach, while for de novo interface design, a comparison of computational protocols highlights the power and increased experimental success achieved using the machine learning program ProteinMPNN. In relating information from docking and design, we observe that agreement between fragment-based sequence preferences and ProteinMPNN sequence inference correlates with experimental success. Additional insights for designing polar interactions are highlighted by experimentally testing larger and more polar interfaces. In all, using X-ray crystallography and cryo-EM, we report five structures for seven protein cages, with atomic resolution in the best case reaching 2.0 Å. We also report structures of two incompletely assembled protein cages, providing unique insights into one type of assembly failure. The new set of designed cages and their structures add substantially to the body of available protein nanoparticles, and to methodologies for their creation.

Critical Assessment of Methods for Predicting the 3D Structure of Proteins and Protein Complexes

Advances in a scientific discipline are often measured by small, incremental steps. In this review, we report on two intertwined disciplines in the protein structure prediction field, modeling of single chains and modeling of complexes, that have over decades emulated this pattern, as monitored by the community-wide blind prediction experiments CASP and CAPRI. However, over the past few years, dramatic advances were observed for the accurate prediction of single protein chains, driven by a surge of deep learning methodologies entering the prediction field. We review the mainscientific developments that enabled these recent breakthroughs and feature the important role of blind prediction experiments in building up and nurturing the structure prediction field. We discuss how the new wave of artificial intelligence-based methods is impacting the fields of computational and experimental structural biology and highlight areas in which deep learning methods are likely to lead to future developments, provided that major challenges are overcome.

An artificial neural network model to predict structure-based protein-protein free energy of binding from Rosetta-calculated properties

The prediction of the free energy (ΔG) of binding for protein-protein complexes is of general scientific interest as it has a variety of applications in the fields of molecular and chemical biology, materials science, and biotechnology. Despite its centrality in understanding protein association phenomena and protein engineering, the ΔG of binding is a daunting quantity to obtain theoretically. In this work, we devise a novel Artificial Neural Network (ANN) model to predict the ΔG of binding for a given three-dimensional structure of a protein-protein complex with Rosetta-calculated properties. Our model was tested using two data sets, and it presented a root-mean-square error ranging from 1.67 kcal mol^-1 to 2.45 kcal mol^-1, showing a better performance compared to the available state-of-the-art tools. Validation of the model for a variety of protein-protein complexes is showcased.

Protein-protein interaction prediction methods: from docking-based to AI-based approaches

Protein-protein interactions (PPIs), such as protein-protein inhibitor, antibody-antigen complex, and supercomplexes play diverse and important roles in cells. Recent advances in structural analysis methods, including cryo-EM, for the determination of protein complex structures are remarkable. Nevertheless, much room remains for improvement and utilization of computational methods to predict PPIs because of the large number and great diversity of unresolved complex structures. This review introduces a wide array of computational methods, including our own, for estimating PPIs including antibody-antigen interactions, offering both historical and forward-looking perspectives.

Spatial organization of hydrophobic and charged residues affects protein thermal stability and binding affinity

What are the molecular determinants of protein–protein binding affinity and whether they are similar to those regulating fold stability are two major questions of molecular biology, whose answers bring important implications both from a theoretical and applicative point of view. Here, we analyze chemical and physical features on a large dataset of protein–protein complexes with reliable experimental binding affinity data and compare them with a set of monomeric proteins for which melting temperature data was available. In particular, we probed the spatial organization of protein (1) intramolecular and intermolecular interaction energies among residues, (2) amino acidic composition, and (3) their hydropathy features. Analyzing the interaction energies, we found that strong Coulombic interactions are preferentially associated with a high protein thermal stability, while strong intermolecular van der Waals energies correlate with stronger protein–protein binding affinity. Statistical analysis of amino acids abundances, exposed to the molecular surface and/or in interaction with the molecular partner, confirmed that hydrophobic residues present on the protein surfaces are preferentially located in the binding regions, while charged residues behave oppositely. Leveraging on the important role of van der Waals interface interactions in binding affinity, we focused on the molecular surfaces in the binding regions and evaluated their shape complementarity, decomposing the molecular patches in the 2D Zernike basis. For the first time, we quantified the correlation between local shape complementarity and binding affinity via the Zernike formalism. In addition, considering the solvent interactions via the residue hydropathy, we found that the hydrophobicity of the binding regions dictates their shape complementary as much as the correlation between van der Waals energy and binding affinity. In turn, these relationships pave the way to the fast and accurate prediction and design of optimal binding regions as the 2D Zernike formalism allows a rapid and superposition-free comparison between possible binding surfaces.

Polydopamine nanospheres coated with bovine serum albumin permit enhanced cell differentiation: fundamental mechanism and practical application for protein coating formation

Protein coating is a strategy for modifying and improving the surface functional properties of nanomaterials. However, the underlying mechanism behind protein coating formation, which is essential for its practical applications, remains largely unknown. Herein, we investigate the fundamental molecular mechanism of protein coating formation. Polydopamine nanospheres (PDANS) coated with bovine serum albumin (BSA) are examined in this study due to their wide biomedical potential. Our results demonstrate that BSAs can flexibly bind to PDANS and maintain their structural dynamicity. Our findings unveil that regular structure formation arises from BSAs lateral interactions via electrostatic forces. Notably, the protein coating modified PDANS surface enhances cell adhesion and proliferation as well as osteogenic differentiation. Such an enhancement is attributed to complementary surface properties provided by the dynamic PDANS-BSA complex and regular structure caused by BSA-BSA interactions in protein coating formation. This study provides a fundamental understanding of the molecular mechanism of protein coating formation, which facilitates the further development of functional protein-coated nanomaterials and guides the bioengineering decision making for biomedical applications, especially in bone tissue engineering.

Prediction of protein assemblies, the next frontier: The CASP14-CAPRI experiment

We present the results for CAPRI Round 50, the fourth joint CASP-CAPRI protein assembly prediction challenge. The Round comprised a total of twelve targets, including six dimers, three trimers, and three higher-order oligomers. Four of these were easy targets, for which good structural templates were available either for the full assembly, or for the main interfaces (of the higher-order oligomers). Eight were difficult targets for which only distantly related templates were found for the individual subunits. Twenty-five CAPRI groups including eight automatic servers submitted ~1250 models per target. Twenty groups including six servers participated in the CAPRI scoring challenge submitted ~190 models per target. The accuracy of the predicted models was evaluated using the classical CAPRI criteria. The prediction performance was measured by a weighted scoring scheme that takes into account the number of models of acceptable quality or higher submitted by each group as part of their five top-ranking models. Compared to the previous CASP-CAPRI challenge, top performing groups submitted such models for a larger fraction (70-75%) of the targets in this Round, but fewer of these models were of high accuracy. Scorer groups achieved stronger performance with more groups submitting correct models for 70-80% of the targets or achieving high accuracy predictions. Servers performed less well in general, except for the MDOCKPP and LZERD servers, who performed on par with human groups. In addition to these results, major advances in methodology are discussed, providing an informative overview of where the prediction of protein assemblies currently stands.

Protein-Protein Docking: Past, Present, and Future

The biological significance of proteins attracted the scientific community in exploring their characteristics. The studies shed light on the interaction patterns and functions of proteins in a living body. Due to their practical difficulties, reliable experimental techniques pave the way for introducing computational methods in the interaction prediction. Automated methods reduced the difficulties but could not yet replace experimental studies as the field is still evolving. Interaction prediction problem being critical needs highly accurate results, but none of the existing methods could offer reliable performance that can parallel with experimental results yet. This article aims to assess the existing computational docking algorithms, their challenges, and future scope. Blind docking techniques are quite helpful when no information other than the individual structures are available. As more and more complex structures are being added to different databases, information-driven approaches can be a good alternative. Artificial intelligence, ruling over the major fields, is expected to take over this domain very shortly.

Using diverse potentials and scoring functions for the development of improved machine-learned models for protein-ligand affinity and docking pose prediction

The advent of computational drug discovery holds the promise of significantly reducing the effort of experimentalists, along with monetary cost. More generally, predicting the binding of small organic molecules to biological macromolecules has far-reaching implications for a range of problems, including metabolomics. However, problems such as predicting the bound structure of a protein-ligand complex along with its affinity have proven to be an enormous challenge. In recent years, machine learning-based methods have proven to be more accurate than older methods, many based on simple linear regression. Nonetheless, there remains room for improvement, as these methods are often trained on a small set of features, with a single functional form for any given physical effect, and often with little mention of the rationale behind choosing one functional form over another. Moreover, it is not entirely clear why one machine learning method is favored over another. In this work, we endeavor to undertake a comprehensive effort towards developing high-accuracy, machine-learned scoring functions, systematically investigating the effects of machine learning method and choice of features, and, when possible, providing insights into the relevant physics using methods that assess feature importance. Here, we show synergism among disparate features, yielding adjusted R² with experimental binding affinities of up to 0.871 on an independent test set and enrichment for native bound structures of up to 0.913. When purely physical terms that model enthalpic and entropic effects are used in the training, we use feature importance assessments to probe the relevant physics and hopefully guide future investigators working on this and other computational chemistry problems.

Community-Wide Experimental Evaluation of the PROSS Stability-Design Method

Recent years have seen a dramatic improvement in protein-design methodology. Nevertheless, most methods demand expert intervention, limiting their widespread adoption. By contrast, the PROSS algorithm for improving protein stability and heterologous expression levels has been successfully applied to a range of challenging enzymes and binding proteins. Here, we benchmark the application of PROSS as a stand-alone tool for protein scientists with no or limited experience in modeling. Twelve laboratories from the Protein Production and Purification Partnership in Europe (P4EU) challenged the PROSS algorithm with 14 unrelated protein targets without support from the PROSS developers. For each target, up to six designs were evaluated for expression levels and in some cases, for thermal stability and activity. In nine targets, designs exhibited increased heterologous expression levels either in prokaryotic and/or eukaryotic expression systems under experimental conditions that were tailored for each target protein. Furthermore, we observed increased thermal stability in nine of ten tested targets. In two prime examples, the human Stem Cell Factor (hSCF) and human Cadherin-Like Domain (CLD12) from the RET receptor, the wild type proteins were not expressible as soluble proteins in E. coli, yet the PROSS designs exhibited high expression levels in E. coli and HEK293 cells, respectively, and improved thermal stability. We conclude that PROSS may improve stability and expressibility in diverse cases, and that improvement typically requires target-specific expression conditions. This study demonstrates the strengths of community-wide efforts to probe the generality of new methods and recommends areas for future research to advance practically useful algorithms for protein science.

The Electrostatic Features of Dengue Virus Capsid Assembly

Dengue virus causes serious diseases and deaths in the world. Understanding the fundamental mechanisms of dengue virus is highly demanded to develop treatments for dengue virus caused diseases. Here, we present a computational work which focused on the stability of dengue viral capsid. The interactions among E proteins on the dengue viral capsid were studied using several computational approaches. It was found that the electrostatic distribution on a single E protein monomer is highly inhomogeneous, which makes an E protein strongly binding with another E protein. This is the reason why all the E proteins form homodimers as the basic units on the whole dengue viral capsids. The pKa calculations of E proteins demonstrated that the folding energy of an E protein is low and stable in the range of pH 6–10, which is different from many other proteins that are stable at certain pH. The pH dependence of binding energy of E protein homodimer shows that the binding energy is low and independent from pH when the pH is also in the range of 6–10. This finding implies that the dengue virus can survive in a wide range of pH, which can explain why the dengue virus is so widely distributed in the world and spreads fast.

Residue-based pharmacophore approaches to study protein-protein interactions

This review focuses on pharmacophore approaches in researching protein interfaces that bind protein ligands. Pharmacophore descriptions of binding interfaces that employ molecular dynamics simulation can account for effects of solvation and conformational flexibility. In addition, these calculations provide an approximation to entropic considerations and as such, a better approximation of the free energy of binding. Residue-based pharmacophore approaches can facilitate a variety of drug discovery tasks such as the identification of receptor-ligand partners, identifying their binding poses, designing protein interfaces for selectivity, or defining a reduced mutational combinatorial exploration for subsequent experimental engineering techniques by orders of magnitudes.

Engineered receptors for human cytomegalovirus that are orthogonal to normal human biology

A trimeric glycoprotein complex on the surface of human cytomegalovirus (HCMV) binds to platelet-derived growth factor (PDGF) receptor α (PDGFRα) to mediate host cell recognition and fusion of the viral and cellular membranes. Soluble PDGFRα potently neutralizes HCMV in tissue culture, and its potential use as an antiviral therapeutic has the benefit that any escape mutants will likely be attenuated. However, PDGFRα binds multiple PDGF ligands in the human body as part of developmental programs in embryogenesis and continuing through adulthood. Any therapies with soluble receptor therefore come with serious efficacy and safety concerns, especially for the treatment of congenital HCMV. Soluble virus receptors that are orthogonal to human biology might resolve these concerns. This engineering problem is solved by deep mutational scanning on the D2-D3 domains of PDGFRα to identify variants that maintain interactions with the HCMV glycoprotein trimer in the presence of competing PDGF ligands. Competition by PDGFs is conformation-dependent, whereas HCMV trimer binding is independent of proper D2-D3 conformation, and many mutations at the receptor-PDGF interface are suitable for functionally separating trimer from PDGF interactions. Purified soluble PDGFRα carrying a targeted mutation succeeded in displaying wild type affinity for HCMV trimer with a simultaneous loss of PDGF binding, and neutralizes trimer-only and trimer/pentamer-expressing HCMV strains infecting fibroblasts or epithelial cells. Overall, this work makes important progress in the realization of soluble HCMV receptors for clinical application.

Human cytomegalovirus (HCMV) causes severe disease in transplant recipients and immunocompromised patients, and infections in a fetus or neonate are responsible for life-long neurological defects. Cell entry is in part mediated by a trimeric glycoprotein complex on the viral surface, which binds tightly to the host receptor PDGFRα. The soluble extracellular region of PDGFRα can be used as an antiviral agent to potently neutralize the virus in vitro. However, PDGFRα ordinarily binds growth factors in the human body to regulate developmental programs, which will limit the in vivo efficacy and safety of soluble PDGFRα. Using saturation mutagenesis and selections in human cell culture, mutations in PDGFRα are identified that eliminate off-target growth factor interactions while preserving HCMV binding and neutralization.

Generating quantitative binding landscapes through fractional binding selections combined with deep sequencing and data normalization

Quantifying the effects of various mutations on binding free energy is crucial for understanding the evolution of protein-protein interactions and would greatly facilitate protein engineering studies. Yet, measuring changes in binding free energy (ΔΔG_bind) remains a tedious task that requires expression of each mutant, its purification, and affinity measurements. We developed an attractive approach that allows us to quantify ΔΔG_bind for thousands of protein mutants in one experiment. Our protocol combines protein randomization, Yeast Surface Display technology, deep sequencing, and a few experimental ΔΔG_bind data points on purified proteins to generate ΔΔG_bind values for the remaining numerous mutants of the same protein complex. Using this methodology, we comprehensively map the single-mutant binding landscape of one of the highest-affinity interaction between BPTI and Bovine Trypsin (BT). We show that ΔΔG_bind for this interaction could be quantified with high accuracy over the range of 12 kcal mol⁻¹ displayed by various BPTI single mutants.

Quantifying the effect of mutations on binding free energy is important to understand protein-protein interaction (PPI). Here the authors develop a method based on yeast display and next-generation sequencing to generate quantitative binding landscapes for any PPI regardless of their Kd value.

Blind prediction of homo- and hetero-protein complexes: The CASP13-CAPRI experiment

We present the results for CAPRI Round 46, the third joint CASP-CAPRI protein assembly prediction challenge. The Round comprised a total of 20 targets including 14 homo-oligomers and 6 heterocomplexes. Eight of the homo-oligomer targets and one heterodimer comprised proteins that could be readily modeled using templates from the Protein Data Bank, often available for the full assembly. The remaining 11 targets comprised 5 homodimers, 3 heterodimers, and two higher-order assemblies. These were more difficult to model, as their prediction mainly involved "ab-initio" docking of subunit models derived from distantly related templates. A total of ~30 CAPRI groups, including 9 automatic servers, submitted on average ~2000 models per target. About 17 groups participated in the CAPRI scoring rounds, offered for most targets, submitting ~170 models per target. The prediction performance, measured by the fraction of models of acceptable quality or higher submitted across all predictors groups, was very good to excellent for the nine easy targets. Poorer performance was achieved by predictors for the 11 difficult targets, with medium and high quality models submitted for only 3 of these targets. A similar performance "gap" was displayed by scorer groups, highlighting yet again the unmet challenge of modeling the conformational changes of the protein components that occur upon binding or that must be accounted for in template-based modeling. Our analysis also indicates that residues in binding interfaces were less well predicted in this set of targets than in previous Rounds, providing useful insights for directions of future improvements.

Void distributions reveal structural link between jammed packings and protein cores

Dense packing of hydrophobic residues in the cores of globular proteins determines their stability. Recently, we have shown that protein cores possess packing fraction ϕ≈0.56, which is the same as dense, random packing of amino-acid-shaped particles. In this article, we compare the structural properties of protein cores and jammed packings of amino-acid-shaped particles in much greater depth by measuring their local and connected void regions. We find that the distributions of surface Voronoi cell volumes and local porosities obey similar statistics in both systems. We also measure the probability that accessible, connected void regions percolate as a function of the size of a spherical probe particle and show that both systems possess the same critical probe size. We measure the critical exponent τ that characterizes the size distribution of connected void clusters at the onset of percolation. We find that the cluster size statistics are similar for void percolation in packings of amino-acid-shaped particles and randomly placed spheres, but different from that for void percolation in jammed sphere packings. We propose that the connected void regions are a defining structural feature of proteins and can be used to differentiate experimentally observed proteins from decoy structures that are generated using computational protein design software. This work emphasizes that jammed packings of amino-acid-shaped particles can serve as structural and mechanical analogs of protein cores, and could therefore be useful in modeling the response of protein cores to cavity-expanding and -reducing mutations.

Third generation antibody discovery methods: in silico rational design

Owing to their outstanding performances in molecular recognition, antibodies are extensively used in research and applications in molecular biology, biotechnology and medicine. Recent advances in experimental and computational methods are making it possible to complement well-established in vivo (first generation) and in vitro (second generation) methods of antibody discovery with novel in silico (third generation) approaches. Here we describe the principles of computational antibody design and review the state of the art in this field. We then present Modular, a method that implements the rational design of antibodies in a modular manner, and describe the opportunities offered by this approach.

Ultrahigh specificity in a network of computationally designed protein-interaction pairs

Protein networks in all organisms comprise homologous interacting pairs. In these networks, some proteins are specific, interacting with one or a few binding partners, whereas others are multispecific and bind a range of targets. We describe an algorithm that starts from an interacting pair and designs dozens of new pairs with diverse backbone conformations at the binding site as well as new binding orientations and sequences. Applied to a high-affinity bacterial pair, the algorithm results in 18 new ones, with cognate affinities from pico- to micromolar. Three pairs exhibit 3-5 orders of magnitude switch in specificity relative to the wild type, whereas others are multispecific, collectively forming a protein-interaction network. Crystallographic analysis confirms design accuracy, including in new backbones and polar interactions. Preorganized polar interaction networks are responsible for high specificity, thus defining design principles that can be applied to program synthetic cellular interaction networks of desired affinity and specificity.

The molecular basis of ultrahigh specificity in protein-protein interactions remains obscure. The authors present a computational method to design atomically accurate new pairs exhibiting >100,000-fold specificity switches, generating a large and complex interaction network.

Local Interaction Signal Analysis Predicts Protein-Protein Binding Affinity

Several models estimating the strength of the interaction between proteins in a complex have been proposed. By exploring the geometry of contact distribution at protein-protein interfaces, we provide an improved model of binding energy. Local interaction signal analysis (LISA) is a radial function based on terms describing favorable and non-favorable contacts obtained by density functional theory, the support-core-rim interface residue distribution, non-interacting charged residues and secondary structures contribution. The three-dimensional organization of the contacts and their contribution on localized hot-sites over the entire interaction surface were numerically evaluated. LISA achieves a correlation of 0.81 (and a root-mean-square error of 2.35 ± 0.38 kcal/mol) when tested on 125 complexes for which experimental measurements were realized. LISA's performance is stable for subsets defined by functional composition and extent of conformational changes upon complex formation. A large-scale comparison with 17 other functions demonstrated the power of the geometrical model in the understanding of complex binding.

Comparing side chain packing in soluble proteins, protein-protein interfaces, and transmembrane proteins

We compare side chain prediction and packing of core and non-core regions of soluble proteins, protein-protein interfaces, and transmembrane proteins. We first identified or created comparable databases of high-resolution crystal structures of these 3 protein classes. We show that the solvent-inaccessible cores of the 3 classes of proteins are equally densely packed. As a result, the side chains of core residues at protein-protein interfaces and in the membrane-exposed regions of transmembrane proteins can be predicted by the hard-sphere plus stereochemical constraint model with the same high prediction accuracies (>90%) as core residues in soluble proteins. We also find that for all 3 classes of proteins, as one moves away from the solvent-inaccessible core, the packing fraction decreases as the solvent accessibility increases. However, the side chain predictability remains high (80% within 30°) up to a relative solvent accessibility, rSASA≲0.3, for all 3 protein classes. Our results show that ≈40% of the interface regions in protein complexes are "core", that is, densely packed with side chain conformations that can be accurately predicted using the hard-sphere model. We propose packing fraction as a metric that can be used to distinguish real protein-protein interactions from designed, non-binding, decoys. Our results also show that cores of membrane proteins are the same as cores of soluble proteins. Thus, the computational methods we are developing for the analysis of the effect of hydrophobic core mutations in soluble proteins will be equally applicable to analyses of mutations in membrane proteins.

Antibody Affinity Maturation by Computational Design

The immune systems protect our bodies from foreign molecules or antigens, where antibodies play important roles. Antibodies evolve over time upon antigen encounter by somatically mutating their genome sequences. The end result is a series of antibodies that display higher affinities and specificities to specific antigens. This process is called affinity maturation. Recent improvements in computer hardware and modeling algorithms now enable the rational design of protein structures and functions, and several works on computer-aided antibody design have been published. In this chapter, we briefly describe computational methods for antibody affinity maturation, focusing on methods for sampling antibody conformations and for scoring designed antibody variants. We also discuss lessons learned from the successful computer-aided design of antibodies.

The challenge of modeling protein assemblies: the CASP12-CAPRI experiment

We present the quality assessment of 5613 models submitted by predictor groups from both CAPRI and CASP for the total of 15 most tractable targets from the second joint CASP-CAPRI protein assembly prediction experiment. These targets comprised 12 homo-oligomers and 3 hetero-complexes. The bulk of the analysis focuses on 10 targets (of CAPRI Round 37), which included all 3 hetero-complexes, and whose protein chains or the full assembly could be readily modeled from structural templates in the PDB. On average, 28 CAPRI groups and 10 CASP groups (including automatic servers), submitted models for each of these 10 targets. Additionally, about 16 groups participated in the CAPRI scoring experiments. A range of acceptable to high quality models were obtained for 6 of the 10 Round 37 targets, for which templates were available for the full assembly. Poorer results were achieved for the remaining targets due to the lower quality of the templates available for the full complex or the individual protein chains, highlighting the unmet challenge of modeling the structural adjustments of the protein components that occur upon binding or which must be accounted for in template-based modeling. On the other hand, our analysis indicated that residues in binding interfaces were correctly predicted in a sizable fraction of otherwise poorly modeled assemblies and this with higher accuracy than published methods that do not use information on the binding partner. Lastly, the strengths and weaknesses of the assessment methods are evaluated and improvements suggested.

The computational prediction of protein assemblies

The function of proteins in the cell is almost always mediated by their interaction with different partners, including other proteins, nucleic acids or small organic molecules. The ability of identifying all of them is an essential step in our quest for understanding life at the molecular level. The inference of the protein complex composition and of its molecular details can also provide relevant clues for the development and the design of drugs. In this short review, I will discuss the computational aspects of the analysis and prediction of protein-protein assemblies and discuss some of the most recent developments as seen in the last Critical Assessment of Techniques for Protein Structure Prediction (CASP) experiment.

Structure-based cross-docking analysis of antibody-antigen interactions

Antibody–antigen interactions are critical to our immune response, and understanding the structure-based biophysical determinants for their binding specificity and affinity is of fundamental importance. We present a computational structure-based cross-docking study to test the identification of native antibody–antigen interaction pairs among cognate and non-cognate complexes. We picked a dataset of 17 antibody–antigen complexes of which 11 have both bound and unbound structures available, and we generated a representative ensemble of cognate and non-cognate complexes. Using the Rosetta interface score as a classifier, the cognate pair was the top-ranked model in 80% (14/17) of the antigen targets using bound monomer structures in docking, 35% (6/17) when using unbound, and 12% (2/17) when using the homology-modeled backbones to generate the complexes. Increasing rigid-body diversity of the models using RosettaDock’s local dock routine lowers the discrimination accuracy with the cognate antibody–antigen pair ranking in bound and unbound models but recovers additional top-ranked cognate complexes when using homology models. The study is the first structure-based cross-docking attempt aimed at distinguishing antibody–antigen binders from non-binders and demonstrates the challenges to address for the methods to be widely applicable to supplement high-throughput experimental antibody sequencing workflows.

FlexPepDock lessons from CAPRI peptide-protein rounds and suggested new criteria for assessment of model quality and utility

CAPRI rounds 28 and 29 included, for the first time, peptide-receptor targets of three different systems, reflecting increased appreciation of the importance of peptide-protein interactions. The CAPRI rounds allowed us to objectively assess the performance of Rosetta FlexPepDock, one of the first protocols to explicitly include peptide flexibility in docking, accounting for peptide conformational changes upon binding. We discuss here successes and challenges in modeling these targets: we obtain top-performing, high-resolution models of the peptide motif for cases with known binding sites but there is a need for better modeling of flanking regions, as well as better selection criteria, in particular for unknown binding sites. These rounds have also provided us the opportunity to reassess the success criteria, to better reflect the quality of a peptide-protein complex model. Using all models submitted to CAPRI, we analyze the correlation between current classification criteria and the ability to retrieve critical interface features, such as hydrogen bonds and hotspots. We find that loosening the backbone (and ligand) RMSD threshold, together with a restriction on the side chain RMSD measure, allows us to improve the selection of high-accuracy models. We also suggest a new measure to assess interface hydrogen bond recovery, which is not assessed by the current CAPRI criteria. Finally, we find that surprisingly much can be learned from rather inaccurate models about binding hotspots, suggesting that the current status of peptide-protein docking methods, as reflected by the submitted CAPRI models, can already have a significant impact on our understanding of protein interactions. Proteins 2017; 85:445-462. © 2016 Wiley Periodicals, Inc.

Trade-offs between enzyme fitness and solubility illuminated by deep mutational scanning

Proteins are marginally stable, and an understanding of the sequence determinants for improved protein solubility is highly desired. For enzymes, it is well known that many mutations that increase protein solubility decrease catalytic activity. These competing effects frustrate efforts to design and engineer stable, active enzymes without laborious high-throughput activity screens. To address the trade-off between enzyme solubility and activity, we performed deep mutational scanning using two different screens/selections that purport to gauge protein solubility for two full-length enzymes. We assayed a TEM-1 beta-lactamase variant and levoglucosan kinase (LGK) using yeast surface display (YSD) screening and a twin-arginine translocation pathway selection. We then compared these scans with published experimental fitness landscapes. Results from the YSD screen could explain 37% of the variance in the fitness landscapes for one enzyme. Five percent to 10% of all single missense mutations improve solubility, matching theoretical predictions of global protein stability. For a given solubility-enhancing mutation, the probability that it would retain wild-type fitness was correlated with evolutionary conservation and distance to active site, and anticorrelated with contact number. Hybrid classification models were developed that could predict solubility-enhancing mutations that maintain wild-type fitness with an accuracy of 90%. The downside of using such classification models is the removal of rare mutations that improve both fitness and solubility. To reveal the biophysical basis of enhanced protein solubility and function, we determined the crystallographic structure of one such LGK mutant. Beyond fundamental insights into trade-offs between stability and activity, these results have potential biotechnological applications.

BindML/BindML+: Detecting Protein-Protein Interaction Interface Propensity from Amino Acid Substitution Patterns

Prediction of protein-protein interaction sites in a protein structure provides important information for elucidating the mechanism of protein function and can also be useful in guiding a modeling or design procedures of protein complex structures. Since prediction methods essentially assess the propensity of amino acids that are likely to be part of a protein docking interface, they can help in designing protein-protein interactions. Here, we introduce BindML and BindML+ protein-protein interaction sites prediction methods. BindML predicts protein-protein interaction sites by identifying mutation patterns found in known protein-protein complexes using phylogenetic substitution models. BindML+ is an extension of BindML for distinguishing permanent and transient types of protein-protein interaction sites. We developed an interactive web-server that provides a convenient interface to assist in structural visualization of protein-protein interactions site predictions. The input data for the web-server are a tertiary structure of interest. BindML and BindML+ are available at http://kiharalab.org/bindml/ and http://kiharalab.org/bindml/plus/ .

Elucidating Mechanisms of Molecular Recognition Between Human Argonaute and miRNA Using Computational Approaches

MicroRNA (miRNA) and Argonaute (AGO) protein together form the RNA-induced silencing complex (RISC) that plays an essential role in the regulation of gene expression. Elucidating the underlying mechanism of AGO-miRNA recognition is thus of great importance not only for the in-depth understanding of miRNA function but also for inspiring new drugs targeting miRNAs. In this chapter we introduce a combined computational approach of molecular dynamics (MD) simulations, Markov state models (MSMs), and protein-RNA docking to investigate AGO-miRNA recognition. Constructed from MD simulations, MSMs can elucidate the conformational dynamics of AGO at biologically relevant timescales. Protein-RNA docking can then efficiently identify the AGO conformations that are geometrically accessible to miRNA. Using our recent work on human AGO2 as an example, we explain the rationale and the workflow of our method in details. This combined approach holds great promise to complement experiments in unraveling the mechanisms of molecular recognition between large, flexible, and complex biomolecules.

ProtLID, a Residue-Based Pharmacophore Approach to Identify Cognate Protein Ligands in the Immunoglobulin Superfamily

Members of the extracellular immunoglobulin superfamily (IgSF) play a key role in immune regulation through the control of the co-stimulatory pathway, and have emerged as potent drug targets in cancers, infectious diseases, and autoimmunity. Despite the difficult experimental access to this class of proteins, single structures of ectodomains of IgSF proteins are solved with regularity. However, the most biologically critical challenge for this class of proteins is the identification of their cognate ligands that communicate intercellular signals. We describe a conceptually novel method, protein-ligand interface design (ProtLID), to identify cognate ligands from a subproteome for a given target receptor protein. ProtLID designs an optimal protein interface for a given receptor by running extensive molecular dynamics simulations of single-residue probes. The type and location of residue preferences establish a residue-based pharmacophore, which is subsequently used to find potential matches among candidate ligands within a subproteome.

Modeling protein-protein and protein-peptide complexes: CAPRI 6th edition

We present the sixth report evaluating the performance of methods for predicting the atomic resolution structures of protein complexes offered as targets to the community-wide initiative on the Critical Assessment of Predicted Interactions (CAPRI). The evaluation is based on a total of 20,670 predicted models for 8 protein-peptide complexes, a novel category of targets in CAPRI, and 12 protein-protein targets in CAPRI prediction Rounds held during the years 2013-2016. For two of the protein-protein targets, the focus was on the prediction of side-chain conformation and positions of interfacial water molecules. Seven of the protein-protein targets were particularly challenging owing to their multicomponent nature, to conformational changes at the binding site, or to a combination of both. Encouragingly, the very large multiprotein complex with the nucleosome was correctly predicted, and correct models were submitted for the protein-peptide targets, but not for some of the challenging protein-protein targets. Models of acceptable quality or better were obtained for 14 of the 20 targets, including medium quality models for 13 targets and high quality models for 8 targets, indicating tangible progress of present-day computational methods in modeling protein complexes with increased accuracy. Our evaluation suggests that the progress stems from better integration of different modeling tools with docking procedures, as well as the use of more sophisticated evolutionary information to score models. Nonetheless, adequate modeling of conformational flexibility in interacting proteins remains an important area with a crucial need for improvement. Proteins 2017; 85:359-377. © 2016 Wiley Periodicals, Inc.

A generalized framework for computational design and mutational scanning of T-cell receptor binding interfaces

T-cell receptors (TCRs) have emerged as a new class of therapeutics, most prominently for cancer where they are the key components of new cellular therapies as well as soluble biologics. Many studies have generated high affinity TCRs in order to enhance sensitivity. Recent outcomes, however, have suggested that fine manipulation of TCR binding, with an emphasis on specificity may be more valuable than large affinity increments. Structure-guided design is ideally suited for this role, and here we studied the generality of structure-guided design as applied to TCRs. We found that a previous approach, which successfully optimized the binding of a therapeutic TCR, had poor accuracy when applied to a broader set of TCR interfaces. We thus sought to develop a more general purpose TCR design framework. After assembling a large dataset of experimental data spanning multiple interfaces, we trained a new scoring function that accounted for unique features of each interface. Together with other improvements, such as explicit inclusion of molecular flexibility, this permitted the design new affinity-enhancing mutations in multiple TCRs, including those not used in training. Our approach also captured the impacts of mutations and substitutions in the peptide/MHC ligand, and recapitulated recent findings regarding TCR specificity, indicating utility in more general mutational scanning of TCR-pMHC interfaces.

Protein social behavior makes a stronger signal for partner identification than surface geometry

Cells are interactive living systems where proteins movements, interactions and regulation are substantially free from centralized management. How protein physico‐chemical and geometrical properties determine who interact with whom remains far from fully understood. We show that characterizing how a protein behaves with many potential interactors in a complete cross‐docking study leads to a sharp identification of its cellular/true/native partner(s). We define a sociability index, or S‐index, reflecting whether a protein likes or not to pair with other proteins. Formally, we propose a suitable normalization function that accounts for protein sociability and we combine it with a simple interface‐based (ranking) score to discriminate partners from non‐interactors. We show that sociability is an important factor and that the normalization permits to reach a much higher discriminative power than shape complementarity docking scores. The social effect is also observed with more sophisticated docking algorithms. Docking conformations are evaluated using experimental binding sites. These latter approximate in the best possible way binding sites predictions, which have reached high accuracy in recent years. This makes our analysis helpful for a global understanding of partner identification and for suggesting discriminating strategies. These results contradict previous findings claiming the partner identification problem being solvable solely with geometrical docking. Proteins 2016; 85:137–154. © 2016 Wiley Periodicals, Inc.

Structural Interface Forms and Their Involvement in Stabilization of Multidomain Proteins or Protein Complexes

The presented analysis concerns the inter-domain and inter-protein interface in protein complexes. We propose extending the traditional understanding of the protein domain as a function of local compactness with an additional criterion which refers to the presence of a well-defined hydrophobic core. Interface areas in selected homodimers vary with respect to their contribution to share as well as individual (domain-specific) hydrophobic cores. The basic definition of a protein domain, i.e., a structural unit characterized by tighter packing than its immediate environment, is extended in order to acknowledge the role of a structured hydrophobic core, which includes the interface area. The hydrophobic properties of interfaces vary depending on the status of interacting domains—In this context we can distinguish: (1) Shared hydrophobic cores (spanning the whole dimer); (2) Individual hydrophobic cores present in each monomer irrespective of whether the dimer contains a shared core. Analysis of interfaces in dystrophin and utrophin indicates the presence of an additional quasi-domain with a prominent hydrophobic core, consisting of fragments contributed by both monomers. In addition, we have also attempted to determine the relationship between the type of interface (as categorized above) and the biological function of each complex. This analysis is entirely based on the fuzzy oil drop model.

Deep Mutational Scans as a Guide to Engineering High Affinity T Cell Receptor Interactions with Peptide-bound Major Histocompatibility Complex

Proteins are often engineered to have higher affinity for their ligands to achieve therapeutic benefit. For example, many studies have used phage or yeast display libraries of mutants within complementarity-determining regions to affinity mature antibodies and T cell receptors (TCRs). However, these approaches do not allow rapid assessment or evolution across the entire interface. By combining directed evolution with deep sequencing, it is now possible to generate sequence fitness landscapes that survey the impact of every amino acid substitution across the entire protein-protein interface. Here we used the results of deep mutational scans of a TCR-peptide-MHC interaction to guide mutational strategies. The approach yielded stable TCRs with affinity increases of >200-fold. The substitutions with the greatest enrichments based on the deep sequencing were validated to have higher affinity and could be combined to yield additional improvements. We also conducted in silico binding analyses for every substitution to compare them with the fitness landscape. Computational modeling did not effectively predict the impacts of mutations distal to the interface and did not account for yeast display results that depended on combinations of affinity and protein stability. However, computation accurately predicted affinity changes for mutations within or near the interface, highlighting the complementary strengths of computational modeling and yeast surface display coupled with deep mutational scanning for engineering high affinity TCRs.

Prediction of homoprotein and heteroprotein complexes by protein docking and template-based modeling: A CASP-CAPRI experiment

We present the results for CAPRI Round 30, the first joint CASP-CAPRI experiment, which brought together experts from the protein structure prediction and protein-protein docking communities. The Round comprised 25 targets from amongst those submitted for the CASP11 prediction experiment of 2014. The targets included mostly homodimers, a few homotetramers, and two heterodimers, and comprised protein chains that could readily be modeled using templates from the Protein Data Bank. On average 24 CAPRI groups and 7 CASP groups submitted docking predictions for each target, and 12 CAPRI groups per target participated in the CAPRI scoring experiment. In total more than 9500 models were assessed against the 3D structures of the corresponding target complexes. Results show that the prediction of homodimer assemblies by homology modeling techniques and docking calculations is quite successful for targets featuring large enough subunit interfaces to represent stable associations. Targets with ambiguous or inaccurate oligomeric state assignments, often featuring crystal contact-sized interfaces, represented a confounding factor. For those, a much poorer prediction performance was achieved, while nonetheless often providing helpful clues on the correct oligomeric state of the protein. The prediction performance was very poor for genuine tetrameric targets, where the inaccuracy of the homology-built subunit models and the smaller pair-wise interfaces severely limited the ability to derive the correct assembly mode. Our analysis also shows that docking procedures tend to perform better than standard homology modeling techniques and that highly accurate models of the protein components are not always required to identify their association modes with acceptable accuracy. Proteins 2016; 84(Suppl 1):323-348. © 2016 Wiley Periodicals, Inc.

Accurate design of megadalton-scale two-component icosahedral protein complexes

Nature provides many examples of self- and co-assembling protein-based molecular machines, including icosahedral protein cages that serve as scaffolds, enzymes, and compartments for essential biochemical reactions and icosahedral virus capsids, which encapsidate and protect viral genomes and mediate entry into host cells. Inspired by these natural materials, we report the computational design and experimental characterization of co-assembling, two-component, 120-subunit icosahedral protein nanostructures with molecular weights (1.8 to 2.8 megadaltons) and dimensions (24 to 40 nanometers in diameter) comparable to those of small viral capsids. Electron microscopy, small-angle x-ray scattering, and x-ray crystallography show that 10 designs spanning three distinct icosahedral architectures form materials closely matching the design models. In vitro assembly of icosahedral complexes from independently purified components occurs rapidly, at rates comparable to those of viral capsids, and enables controlled packaging of molecular cargo through charge complementarity. The ability to design megadalton-scale materials with atomic-level accuracy and controllable assembly opens the door to a new generation of genetically programmable protein-based molecular machines.

Protein design: Past, present, and future

Building on the pioneering work of Ho and DeGrado (J Am Chem Soc 1987, 109, 6751-6758) in the late 1980s, protein design approaches have revealed many fundamental features of protein structure and stability. We are now in the era that the early work presaged - the design of new proteins with practical applications and uses. Here we briefly survey some past milestones in protein design, in addition to highlighting recent progress and future aspirations.

Molecular docking as a popular tool in drug design, an in silico travel

New molecular modeling approaches, driven by rapidly improving computational platforms, have allowed many success stories for the use of computer-assisted drug design in the discovery of new mechanism-or structure-based drugs. In this overview, we highlight three aspects of the use of molecular docking. First, we discuss the combination of molecular and quantum mechanics to investigate an unusual enzymatic mechanism of a flavoprotein. Second, we present recent advances in anti-infectious agents' synthesis driven by structural insights. At the end, we focus on larger biological complexes made by protein-protein interactions and discuss their relevance in drug design. This review provides information on how these large systems, even in the presence of the solvent, can be investigated with the outlook of drug discovery.

Design and fine-tuning redox potentials of metalloproteins involved in electron transfer in bioenergetics

Redox potentials are a major contributor in controlling the electron transfer (ET) rates and thus regulating the ET processes in the bioenergetics. To maximize the efficiency of the ET process, one needs to master the art of tuning the redox potential, especially in metalloproteins, as they represent major classes of ET proteins. In this review, we first describe the importance of tuning the redox potential of ET centers and its role in regulating the ET in bioenergetic processes including photosynthesis and respiration. The main focus of this review is to summarize recent work in designing the ET centers, namely cupredoxins, cytochromes, and iron-sulfur proteins, and examples in design of protein networks involved these ET centers. We then discuss the factors that affect redox potentials of these ET centers including metal ion, the ligands to metal center and interactions beyond the primary ligand, especially non-covalent secondary coordination sphere interactions. We provide examples of strategies to fine-tune the redox potential using both natural and unnatural amino acids and native and nonnative cofactors. Several case studies are used to illustrate recent successes in this area. Outlooks for future endeavors are also provided. This article is part of a Special Issue entitled Biodesign for Bioenergetics--the design and engineering of electronic transfer cofactors, proteins and protein networks, edited by Ronald L. Koder and J.L. Ross Anderson.

Principles and Overview of Sampling Methods for Modeling Macromolecular Structure and Dynamics

Investigation of macromolecular structure and dynamics is fundamental to understanding how macromolecules carry out their functions in the cell. Significant advances have been made toward this end in silico, with a growing number of computational methods proposed yearly to study and simulate various aspects of macromolecular structure and dynamics. This review aims to provide an overview of recent advances, focusing primarily on methods proposed for exploring the structure space of macromolecules in isolation and in assemblies for the purpose of characterizing equilibrium structure and dynamics. In addition to surveying recent applications that showcase current capabilities of computational methods, this review highlights state-of-the-art algorithmic techniques proposed to overcome challenges posed in silico by the disparate spatial and time scales accessed by dynamic macromolecules. This review is not meant to be exhaustive, as such an endeavor is impossible, but rather aims to balance breadth and depth of strategies for modeling macromolecular structure and dynamics for a broad audience of novices and experts.

AB-Bind: Antibody binding mutational database for computational affinity predictions

Antibodies (Abs) are a crucial component of the immune system and are often used as diagnostic and therapeutic agents. The need for high-affinity and high-specificity antibodies in research and medicine is driving the development of computational tools for accelerating antibody design and discovery. We report a diverse set of antibody binding data with accompanying structures that can be used to evaluate methods for modeling antibody interactions. Our Antibody-Bind (AB-Bind) database includes 1101 mutants with experimentally determined changes in binding free energies (ΔΔG) across 32 complexes. Using the AB-Bind data set, we evaluated the performance of protein scoring potentials in their ability to predict changes in binding free energies upon mutagenesis. Numerical correlations between computed and observed ΔΔG values were low (r = 0.16-0.45), but the potentials exhibited predictive power for classifying variants as improved vs weakened binders. Performance was evaluated using the area under the curve (AUC) for receiver operator characteristic (ROC) curves; the highest AUC values for 527 mutants with |ΔΔG| > 1.0 kcal/mol were 0.81, 0.87, and 0.88 using STATIUM, FoldX, and Discovery Studio scoring potentials, respectively. Some methods could also enrich for variants with improved binding affinity; FoldX and Discovery Studio were able to correctly rank 42% and 30%, respectively, of the 80 most improved binders (those with ΔΔG < -1.0 kcal/mol) in the top 5% of the database. This modest predictive performance has value but demonstrates the continuing need to develop and improve protein energy functions for affinity prediction.