SKEMPI 2.0: an updated benchmark of changes in protein-protein binding energy, kinetics and thermodynamics upon mutation

Systematic investigation of machine learning on limited data: A study on predicting protein-protein binding strength

The application of machine learning techniques in biological research, especially when dealing with limited data availability, poses significant challenges. In this study, we leveraged advancements in method development for predicting protein-protein binding strength to conduct a systematic investigation into the application of machine learning on limited data. The binding strength, quantitatively measured as binding affinity, is vital for understanding the processes of recognition, association, and dysfunction that occur within protein complexes. By incorporating transfer learning, integrating domain knowledge, and employing both deep learning and traditional machine learning algorithms, we mitigated the impact of data limitations and made significant advancements in predicting protein-protein binding affinity. In particular, we developed over 20 models, ultimately selecting three representative best-performing ones that belong to distinct categories. The first model is structure-based, consisting of a random forest regression and thirteen handcrafted features. The second model is sequence-based, employing an architecture that combines transferred embedding features with a multilayer perceptron. Finally, we created an ensemble model by averaging the predictions of the two aforementioned models. The comparison with other predictors on three independent datasets confirms the significant improvements achieved by our models in predicting protein-protein binding affinity. The programs for running these three models are available at https://github.com/minghuilab/BindPPI.

A benchmark for evaluation of structure-based online tools for antibody-antigen binding affinity

The prediction of binding affinity changes caused by missense mutations can elucidate antigen-antibody interactions. A few accessible structure-based online computational tools have been proposed. However, selecting suitable software for particular research is challenging, especially research on the SARS-CoV-2 spike protein with antibodies. Therefore, benchmarking of the mutation-diverse SARS-CoV-2 datasets is critical. Here, we collected the datasets including 1216 variants about the changes in binding affinity of antigens from 22 complexes for SARS-CoV-2 S proteins and 22 monoclonal antibodies as well as applied them to evaluate the performance of seven binding affinity prediction tools. The tested tools' Pearson correlations between predicted and measured changes in binding affinity were between -0.158 and 0.657, while accuracy in classification tasks on predicting increasing or decreasing affinity ranged from 0.444 to 0.834. These tools performed relatively better on predicting single mutations, especially at epitope sites, whereas poor performance on extremely decreasing affinity. The tested tools were relatively insensitive to the experimental techniques used to obtain structures of complexes. In summary, we constructed a list of datasets and evaluated a range of structure-based online prediction tools that will explicate relevant processes of antigen-antibody interactions and enhance the computational design of therapeutic monoclonal antibodies.

DDMut-PPI: predicting effects of mutations on protein-protein interactions using graph-based deep learning

Protein-protein interactions (PPIs) play a vital role in cellular functions and are essential for therapeutic development and understanding diseases. However, current predictive tools often struggle to balance efficiency and precision in predicting the effects of mutations on these complex interactions. To address this, we present DDMut-PPI, a deep learning model that efficiently and accurately predicts changes in PPI binding free energy upon single and multiple point mutations. Building on the robust Siamese network architecture with graph-based signatures from our prior work, DDMut, the DDMut-PPI model was enhanced with a graph convolutional network operated on the protein interaction interface. We used residue-specific embeddings from ProtT5 protein language model as node features, and a variety of molecular interactions as edge features. By integrating evolutionary context with spatial information, this framework enables DDMut-PPI to achieve a robust Pearson correlation of up to 0.75 (root mean squared error: 1.33 kcal/mol) in our evaluations, outperforming most existing methods. Importantly, the model demonstrated consistent performance across mutations that increase or decrease binding affinity. DDMut-PPI offers a significant advancement in the field and will serve as a valuable tool for researchers probing the complexities of protein interactions. DDMut-PPI is freely available as a web server and an application programming interface at https://biosig.lab.uq.edu.au/ddmut_ppi.

Accurate prediction of antibody function and structure using bio-inspired antibody language model

In recent decades, antibodies have emerged as indispensable therapeutics for combating diseases, particularly viral infections. However, their development has been hindered by limited structural information and labor-intensive engineering processes. Fortunately, significant advancements in deep learning methods have facilitated the precise prediction of protein structure and function by leveraging co-evolution information from homologous proteins. Despite these advances, predicting the conformation of antibodies remains challenging due to their unique evolution and the high flexibility of their antigen-binding regions. Here, to address this challenge, we present the Bio-inspired Antibody Language Model (BALM). This model is trained on a vast dataset comprising 336 million 40% nonredundant unlabeled antibody sequences, capturing both unique and conserved properties specific to antibodies. Notably, BALM showcases exceptional performance across four antigen-binding prediction tasks. Moreover, we introduce BALMFold, an end-to-end method derived from BALM, capable of swiftly predicting full atomic antibody structures from individual sequences. Remarkably, BALMFold outperforms those well-established methods like AlphaFold2, IgFold, ESMFold and OmegaFold in the antibody benchmark, demonstrating significant potential to advance innovative engineering and streamline therapeutic antibody development by reducing the need for unnecessary trials. The BALMFold structure prediction server is freely available at https://beamlab-sh.com/models/BALMFold.

DeepPPAPredMut: deep ensemble method for predicting the binding affinity change in protein-protein complexes upon mutation

MOTIVATION

Protein-protein interactions underpin many cellular processes and their disruption due to mutations can lead to diseases. With the evolution of protein structure prediction methods like AlphaFold2 and the availability of extensive experimental affinity data, there is a pressing need for updated computational tools that can efficiently predict changes in binding affinity caused by mutations in protein-protein complexes.

RESULTS

We developed a deep ensemble model that leverages protein sequences, predicted structure-based features, and protein functional classes to accurately predict the change in binding affinity due to mutations. The model achieved a correlation of 0.97 and a mean absolute error (MAE) of 0.35 kcal/mol on the training dataset, and maintained robust performance on the test set with a correlation of 0.72 and a MAE of 0.83 kcal/mol. Further validation using Leave-One-Out Complex (LOOC) cross-validation exhibited a correlation of 0.83 and a MAE of 0.51 kcal/mol, indicating consistent performance.

AVAILABILITY AND IMPLEMENTATION

https://web.iitm.ac.in/bioinfo2/DeepPPAPredMut/index.html.

Multi-level bioinformatics resources support drug target discovery of protein-protein interactions

Drug discovery often begins with a new target. Protein-protein interactions (PPIs) are crucial to multitudinous cellular processes and offer a promising avenue for drug-target discovery. PPIs are characterized by multi-level complexity: at the protein level, interaction networks can be used to identify potential targets, whereas at the residue level, the details of the interactions of individual PPIs can be used to examine a target's druggability. Much great progress has been made in target discovery through multi-level PPI-related computational approaches, but these resources have not been fully discussed. Here, we systematically survey bioinformatics tools for identifying and assessing potential drug targets, examining their characteristics, limitations and applications. This work will aid the integration of the broader protein-to-network context with the analysis of detailed binding mechanisms to support the discovery of drug targets.

Mutational biases favor complexity increases in protein interaction networks after gene duplication

Biological systems can gain complexity over time. While some of these transitions are likely driven by natural selection, the extent to which they occur without providing an adaptive benefit is unknown. At the molecular level, one example is heteromeric complexes replacing homomeric ones following gene duplication. Here, we build a biophysical model and simulate the evolution of homodimers and heterodimers following gene duplication using distributions of mutational effects inferred from available protein structures. We keep the specific activity of each dimer identical, so their concentrations drift neutrally without new functions. We show that for more than 60% of tested dimer structures, the relative concentration of the heteromer increases over time due to mutational biases that favor the heterodimer. However, allowing mutational effects on synthesis rates and differences in the specific activity of homo- and heterodimers can limit or reverse the observed bias toward heterodimers. Our results show that the accumulation of more complex protein quaternary structures is likely under neutral evolution, and that natural selection would be needed to reverse this tendency.

Robust prediction of relative binding energies for protein-protein complex mutations using free energy perturbation calculations

Computational free energy-based methods have the potential to significantly improve throughput and decrease costs of protein design efforts. Such methods must reach a high level of reliability, accuracy, and automation to be effectively deployed in practical industrial settings in a way that impacts protein design projects. Here, we present a benchmark study for the calculation of relative changes in protein-protein binding affinity for single point mutations across a variety of systems from the literature, using free energy perturbation (FEP+) calculations. We describe a method for robust treatment of alternate protonation states for titratable amino acids, which yields improved correlation with and reduced error compared to experimental binding free energies. Following careful analysis of the largest outlier cases in our dataset, we assess limitations of the default FEP+ protocols and introduce an automated script which identifies probable outlier cases that may require additional scrutiny and calculates an empirical correction for a subset of charge-related outliers. Through a series of three additional case study systems, we discuss how protein FEP+ can be applied to real-world protein design projects, and suggest areas of further study.

Computational Approaches to Predict Protein-Protein Interactions in Crowded Cellular Environments

Investigating protein-protein interactions is crucial for understanding cellular biological processes because proteins often function within molecular complexes rather than in isolation. While experimental and computational methods have provided valuable insights into these interactions, they often overlook a critical factor: the crowded cellular environment. This environment significantly impacts protein behavior, including structural stability, diffusion, and ultimately the nature of binding. In this review, we discuss theoretical and computational approaches that allow the modeling of biological systems to guide and complement experiments and can thus significantly advance the investigation, and possibly the predictions, of protein-protein interactions in the crowded environment of cell cytoplasm. We explore topics such as statistical mechanics for lattice simulations, hydrodynamic interactions, diffusion processes in high-viscosity environments, and several methods based on molecular dynamics simulations. By synergistically leveraging methods from biophysics and computational biology, we review the state of the art of computational methods to study the impact of molecular crowding on protein-protein interactions and discuss its potential revolutionizing effects on the characterization of the human interactome.

Computational Mutagenesis of Antibody Fragments: Disentangling Side Chains from ΔΔG Predictions

The development of highly potent antibodies and antibody fragments as binding agents holds significant implications in fields such as biosensing and biotherapeutics. Their binding strength is intricately linked to the arrangement and composition of residues at the binding interface. Computational techniques offer a robust means to predict the three-dimensional structure of these complexes and to assess the affinity changes resulting from mutations. Given the interdependence of structure and affinity prediction, our objective here is to disentangle their roles. We aim to evaluate independently six side-chain reconstruction methods and ten binding affinity estimation techniques. This evaluation was pivotal in predicting affinity alterations due to single mutations, a key step in computational affinity maturation protocols. Our analysis focuses on a data set comprising 27 distinct antibody/hen egg white lysozyme complexes, each with crystal structures and experimentally determined binding affinities. Using six different side-chain reconstruction methods, we transformed each structure into its corresponding mutant via in silico single-point mutations. Subsequently, these structures undergo minimization and molecular dynamics simulation. We therefore estimate ΔΔG values based on the original crystal structure, its energy-minimized form, and the ensuing molecular dynamics trajectories. Our research underscores the critical importance of selecting reliable side-chain reconstruction methods and conducting thorough molecular dynamics simulations to accurately predict the impact of mutations. In summary, our study demonstrates that the integration of conformational sampling and scoring is a potent approach to precisely characterizing mutation processes in single-point mutagenesis protocols and crucial for computational antibody design.

Using protein language models for protein interaction hot spot prediction with limited data

BACKGROUND

Protein language models, inspired by the success of large language models in deciphering human language, have emerged as powerful tools for unraveling the intricate code of life inscribed within protein sequences. They have gained significant attention for their promising applications across various areas, including the sequence-based prediction of secondary and tertiary protein structure, the discovery of new functional protein sequences/folds, and the assessment of mutational impact on protein fitness. However, their utility in learning to predict protein residue properties based on scant datasets, such as protein-protein interaction (PPI)-hotspots whose mutations significantly impair PPIs, remained unclear. Here, we explore the feasibility of using protein language-learned representations as features for machine learning to predict PPI-hotspots using a dataset containing 414 experimentally confirmed PPI-hotspots and 504 PPI-nonhot spots.

RESULTS

Our findings showcase the capacity of unsupervised learning with protein language models in capturing critical functional attributes of protein residues derived from the evolutionary information encoded within amino acid sequences. We show that methods relying on protein language models can compete with methods employing sequence and structure-based features to predict PPI-hotspots from the free protein structure. We observed an optimal number of features for model precision, suggesting a balance between information and overfitting.

CONCLUSIONS

This study underscores the potential of transformer-based protein language models to extract critical knowledge from sparse datasets, exemplified here by the challenging realm of predicting PPI-hotspots. These models offer a cost-effective and time-efficient alternative to traditional experimental methods for predicting certain residue properties. However, the challenge of explaining why specific features are important for determining certain residue properties remains.

DVA: predicting the functional impact of single nucleotide missense variants

BACKGROUND

In the past decade, single nucleotide variants (SNVs) have been identified as having a significant relationship with the development and treatment of diseases. Among them, prioritizing missense variants for further functional impact investigation is an essential challenge in the study of common disease and cancer. Although several computational methods have been developed to predict the functional impacts of variants, the predictive ability of these methods is still insufficient in the Mendelian and cancer missense variants.

RESULTS

We present a novel prediction method called the disease-related variant annotation (DVA) method that predicts the effect of missense variants based on a comprehensive feature set of variants, notably, the allele frequency and protein-protein interaction network feature based on graph embedding. Benchmarked against datasets of single nucleotide missense variants, the DVA method outperforms the state-of-the-art methods by up to 0.473 in the area under receiver operating characteristic curve. The results demonstrate that the proposed method can accurately predict the functional impact of single nucleotide missense variants and substantially outperforms existing methods.

CONCLUSIONS

DVA is an effective framework for identifying the functional impact of disease missense variants based on a comprehensive feature set. Based on different datasets, DVA shows its generalization ability and robustness, and it also provides innovative ideas for the study of the functional mechanism and impact of SNVs.

Leveraging Artificial Intelligence to Expedite Antibody Design and Enhance Antibody-Antigen Interactions

This perspective sheds light on the transformative impact of recent computational advancements in the field of protein therapeutics, with a particular focus on the design and development of antibodies. Cutting-edge computational methods have revolutionized our understanding of protein-protein interactions (PPIs), enhancing the efficacy of protein therapeutics in preclinical and clinical settings. Central to these advancements is the application of machine learning and deep learning, which offers unprecedented insights into the intricate mechanisms of PPIs and facilitates precise control over protein functions. Despite these advancements, the complex structural nuances of antibodies pose ongoing challenges in their design and optimization. Our review provides a comprehensive exploration of the latest deep learning approaches, including language models and diffusion techniques, and their role in surmounting these challenges. We also present a critical analysis of these methods, offering insights to drive further progress in this rapidly evolving field. The paper includes practical recommendations for the application of these computational techniques, supplemented with independent benchmark studies. These studies focus on key performance metrics such as accuracy and the ease of program execution, providing a valuable resource for researchers engaged in antibody design and development. Through this detailed perspective, we aim to contribute to the advancement of antibody design, equipping researchers with the tools and knowledge to navigate the complexities of this field.

Geometric Graph Learning to Predict Changes in Binding Free Energy and Protein Thermodynamic Stability upon Mutation

Accurate prediction of binding free energy changes upon mutations is vital for optimizing drugs, designing proteins, understanding genetic diseases, and cost-effective virtual screening. While machine learning methods show promise in this domain, achieving accuracy and generalization across diverse data sets remains a challenge. This study introduces Geometric Graph Learning for Protein-Protein Interactions (GGL-PPI), a novel approach integrating geometric graph representation and machine learning to forecast mutation-induced binding free energy changes. GGL-PPI leverages atom-level graph coloring and multiscale weighted colored geometric subgraphs to capture structural features of biomolecules, demonstrating superior performance on three standard data sets, namely, AB-Bind, SKEMPI 1.0, and SKEMPI 2.0 data sets. The model's efficacy extends to predicting protein thermodynamic stability in a blind test set, providing unbiased predictions for both direct and reverse mutations and showcasing notable generalization. GGL-PPI's precision in predicting changes in binding free energy and stability due to mutations enhances our comprehension of protein complexes, offering valuable insights for drug design endeavors.

Quantification of biases in predictions of protein-protein binding affinity changes upon mutations

Understanding the impact of mutations on protein-protein binding affinity is a key objective for a wide range of biotechnological applications and for shedding light on disease-causing mutations, which are often located at protein-protein interfaces. Over the past decade, many computational methods using physics-based and/or machine learning approaches have been developed to predict how protein binding affinity changes upon mutations. They all claim to achieve astonishing accuracy on both training and test sets, with performances on standard benchmarks such as SKEMPI 2.0 that seem overly optimistic. Here we benchmarked eight well-known and well-used predictors and identified their biases and dataset dependencies, using not only SKEMPI 2.0 as a test set but also deep mutagenesis data on the severe acute respiratory syndrome coronavirus 2 spike protein in complex with the human angiotensin-converting enzyme 2. We showed that, even though most of the tested methods reach a significant degree of robustness and accuracy, they suffer from limited generalizability properties and struggle to predict unseen mutations. Interestingly, the generalizability problems are more severe for pure machine learning approaches, while physics-based methods are less affected by this issue. Moreover, undesirable prediction biases toward specific mutation properties, the most marked being toward destabilizing mutations, are also observed and should be carefully considered by method developers. We conclude from our analyses that there is room for improvement in the prediction models and suggest ways to check, assess and improve their generalizability and robustness.

Structural coverage of the human interactome

Complex biological processes in cells are embedded in the interactome, representing the complete set of protein-protein interactions. Mapping and analyzing the protein structures are essential to fully comprehending these processes' molecular details. Therefore, knowing the structural coverage of the interactome is important to show the current limitations. Structural modeling of protein-protein interactions requires accurate protein structures. In this study, we mapped all experimental structures to the reference human proteome. Later, we found the enrichment in structural coverage when complementary methods such as homology modeling and deep learning (AlphaFold) were included. We then collected the interactions from the literature and databases to form the reference human interactome, resulting in 117 897 non-redundant interactions. When we analyzed the structural coverage of the interactome, we found that the number of experimentally determined protein complex structures is scarce, corresponding to 3.95% of all binary interactions. We also analyzed known and modeled structures to potentially construct the structural interactome with a docking method. Our analysis showed that 12.97% of the interactions from HuRI and 73.62% and 32.94% from the filtered versions of STRING and HIPPIE could potentially be modeled with high structural coverage or accuracy, respectively. Overall, this paper provides an overview of the current state of structural coverage of the human proteome and interactome.

DG-Affinity: predicting antigen-antibody affinity with language models from sequences

BACKGROUND

Antibody-mediated immune responses play a crucial role in the immune defense of human body. The evolution of bioengineering has led the progress of antibody-derived drugs, showing promising efficacy in cancer and autoimmune disease therapy. A critical step of this development process is obtaining the affinity between antibodies and their binding antigens.

RESULTS

In this study, we introduce a novel sequence-based antigen-antibody affinity prediction method, named DG-Affinity. DG-Affinity uses deep neural networks to efficiently and accurately predict the affinity between antibodies and antigens from sequences, without the need for structural information. The sequences of both the antigen and the antibody are first transformed into embedding vectors by two pre-trained language models, then these embeddings are concatenated into an ConvNeXt framework with a regression task. The results demonstrate the superiority of DG-Affinity over the existing structure-based prediction methods and the sequence-based tools, achieving a Pearson's correlation of over 0.65 on an independent test dataset.

CONCLUSIONS

Compared to the baseline methods, DG-Affinity achieves the best performance and can advance the development of antibody design. It is freely available as an easy-to-use web server at https://www.digitalgeneai.tech/solution/affinity .

Alchemical Calculation of Relative Free Energies for Charge-Changing Mutations at Protein-Protein Interfaces Considering Fixed and Variable Protonation States

The calculation of relative free energies (ΔΔG) for charge-changing mutations at protein-protein interfaces through alchemical methods remains challenging due to variations in the system's net charge during charging steps, the possibility of mutated and contacting ionizable residues occurring in various protonation states, and undersampling issues. In this study, we present a set of strategies, collectively termed TIRST/TIRST-H+, to address some of these challenges. Our approaches combine thermodynamic integration (TI) with the prediction of pKa shifts to calculate ΔΔG values. Moreover, special sets of restraints are employed to keep the alchemically transformed molecules separated. The accuracy of the devised approaches was assessed on a large and diverse data set comprising 164 point mutations of charged residues (Asp, Glu, Lys, and Arg) to Ala at the protein-protein interfaces of complexes with known three-dimensional structures. Mean absolute and root-mean-square errors ranging from 1.38 to 1.66 and 1.89 to 2.44 kcal/mol, respectively, and Pearson correlation coefficients of ∼0.6 were obtained when testing the approaches on the selected data set using the GPU-TI module of Amber18 suite and the ff14SB force field. Furthermore, the inclusion of variable protonation states for the mutated acid residues improved the accuracy of the predicted ΔΔG values. Therefore, our results validate the use of TIRST/TIRST-H+ in prospective studies aimed at evaluating the impact of charge-changing mutations to Ala on the stability of protein-protein complexes.

Enhancing antibody affinity through experimental sampling of non-deleterious CDR mutations predicted by machine learning

The application of machine learning (ML) models to optimize antibody affinity to an antigen is gaining prominence. Unfortunately, the small and biased nature of the publicly available antibody-antigen interaction datasets makes it challenging to build an ML model that can accurately predict binding affinity changes due to mutations (ΔΔG). Recognizing these inherent limitations, we reformulated the problem to ask whether an ML model capable of classifying deleterious vs non-deleterious mutations can guide antibody affinity maturation in a practical setting. To test this hypothesis, we developed a Random Forest classifier (Antibody Random Forest Classifier or AbRFC) with expert-guided features and integrated it into a computational-experimental workflow. AbRFC effectively predicted non-deleterious mutations on an in-house validation dataset that is free of biases seen in the publicly available training datasets. Furthermore, experimental screening of a limited number of predictions from the model (<10^2 designs) identified affinity-enhancing mutations in two unrelated SARS-CoV-2 antibodies, resulting in constructs with up to 1000-fold increased binding to the SARS-COV-2 RBD. Our findings indicate that accurate prediction and screening of non-deleterious mutations using machine learning offers a powerful approach to improving antibody affinity.

Machine Learning-Guided Protein Engineering

Recent progress in engineering highly promising biocatalysts has increasingly involved machine learning methods. These methods leverage existing experimental and simulation data to aid in the discovery and annotation of promising enzymes, as well as in suggesting beneficial mutations for improving known targets. The field of machine learning for protein engineering is gathering steam, driven by recent success stories and notable progress in other areas. It already encompasses ambitious tasks such as understanding and predicting protein structure and function, catalytic efficiency, enantioselectivity, protein dynamics, stability, solubility, aggregation, and more. Nonetheless, the field is still evolving, with many challenges to overcome and questions to address. In this Perspective, we provide an overview of ongoing trends in this domain, highlight recent case studies, and examine the current limitations of machine learning-based methods. We emphasize the crucial importance of thorough experimental validation of emerging models before their use for rational protein design. We present our opinions on the fundamental problems and outline the potential directions for future research.

Deep learning in preclinical antibody drug discovery and development

Antibody drugs have become a key part of biotherapeutics. Patients suffering from various diseases have benefited from antibody therapies. However, its development process is rather long, expensive and risky. To speed up the process, reduce cost and improve success rate, artificial intelligence, especially deep learning methods, have been widely used in all aspects of preclinical antibody drug development, from library generation to hit identification, developability screening, lead selection and optimization. In this review, we systematically summarize antibody encodings, deep learning architectures and models used in preclinical antibody drug discovery and development. We also critically discuss challenges and opportunities, problems and possible solutions, current applications and future directions of deep learning in antibody drug development.

Searching similar local 3D micro-environments in protein structure databases with MicroMiner

The available protein structure data are rapidly increasing. Within these structures, numerous local structural sites depict the details characterizing structure and function. However, searching and analyzing these sites extensively and at scale poses a challenge. We present a new method to search local sites in protein structure databases using residue-defined local 3D micro-environments. We implemented the method in a new tool called MicroMiner and demonstrate the capabilities of residue micro-environment search on the example of structural mutation analysis. Usually, experimental structures for both the wild-type and the mutant are unavailable for comparison. With MicroMiner, we extracted $>255 \times 10^{6}$ amino acid pairs in protein structures from the PDB, exemplifying single mutations' local structural changes for single chains and $>45 \times 10^{6}$ pairs for protein-protein interfaces. We further annotate existing data sets of experimentally measured mutation effects, like $\Delta \Delta G$ measurements, with the extracted structure pairs to combine the mutation effect measurement with the structural change upon mutation. In addition, we show how MicroMiner can bridge the gap between mutation analysis and structure-based drug design tools. MicroMiner is available as a command line tool and interactively on the https://proteins.plus/ webserver.

MpbPPI: a multi-task pre-training-based equivariant approach for the prediction of the effect of amino acid mutations on protein-protein interactions

The accurate prediction of the effect of amino acid mutations for protein-protein interactions (PPI $\Delta \Delta G$) is a crucial task in protein engineering, as it provides insight into the relevant biological processes underpinning protein binding and provides a basis for further drug discovery. In this study, we propose MpbPPI, a novel multi-task pre-training-based geometric equivariance-preserving framework to predict PPI  $\Delta \Delta G$. Pre-training on a strictly screened pre-training dataset is employed to address the scarcity of protein-protein complex structures annotated with PPI $\Delta \Delta G$ values. MpbPPI employs a multi-task pre-training technique, forcing the framework to learn comprehensive backbone and side chain geometric regulations of protein-protein complexes at different scales. After pre-training, MpbPPI can generate high-quality representations capturing the effective geometric characteristics of labeled protein-protein complexes for downstream $\Delta \Delta G$ predictions. MpbPPI serves as a scalable framework supporting different sources of mutant-type (MT) protein-protein complexes for flexible application. Experimental results on four benchmark datasets demonstrate that MpbPPI is a state-of-the-art framework for PPI $\Delta \Delta G$ predictions. The data and source code are available at https://github.com/arantir123/MpbPPI.

Evaluation of Molecular Simulations and Deep Learning Prediction of Antibodies' Recognition of TRBC1 and TRBC2

T cell receptor β-chain constant (TRBC) is a promising class of cancer targets consisting of two highly homologous proteins, TRBC1 and TRBC2. Developing targeted antibody therapeutics against TRBC1 or TRBC2 is expected to eradicate the malignant T cells and preserve half of the normal T cells. Recently, several antibody engineering strategies have been used to modulate the TRBC1 and TRBC2 specificity of antibodies. Here, we used molecular simulation and artificial intelligence methods to quantify the affinity difference in antibodies with various mutations for TRBC1 and TRBC2. The affinity of the existing mutants was verified by FEP calculations aided by the AI. We also performed long-time molecular dynamics simulations to reveal the dynamical antigen recognition mechanisms of the TRBC antibodies.

Inferring the Effects of Protein Variants on Protein-Protein Interactions with Interpretable Transformer Representations

Identifying pathogenetic variants and inferring their impact on protein-protein interactions sheds light on their functional consequences on diseases. Limited by the availability of experimental data on the consequences of protein interaction, most existing methods focus on building models to predict changes in protein binding affinity. Here, we introduced MIPPI, an end-to-end, interpretable transformer-based deep learning model that learns features directly from sequences by leveraging the interaction data from IMEx. MIPPI was specifically trained to determine the types of variant impact (increasing, decreasing, disrupting, and no effect) on protein-protein interactions. We demonstrate the accuracy of MIPPI and provide interpretation through the analysis of learned attention weights, which exhibit correlations with the amino acids interacting with the variant. Moreover, we showed the practicality of MIPPI in prioritizing de novo mutations associated with complex neurodevelopmental disorders and the potential to determine the pathogenic and driving mutations. Finally, we experimentally validated the functional impact of several variants identified in patients with such disorders. Overall, MIPPI emerges as a versatile, robust, and interpretable model, capable of effectively predicting mutation impacts on protein-protein interactions and facilitating the discovery of clinically actionable variants.

Deep-learning-enabled protein-protein interaction analysis for prediction of SARS-CoV-2 infectivity and variant evolution

Host-pathogen interactions and pathogen evolution are underpinned by protein-protein interactions between viral and host proteins. An understanding of how viral variants affect protein-protein binding is important for predicting viral-host interactions, such as the emergence of new pathogenic SARS-CoV-2 variants. Here we propose an artificial intelligence-based framework called UniBind, in which proteins are represented as a graph at the residue and atom levels. UniBind integrates protein three-dimensional structure and binding affinity and is capable of multi-task learning for heterogeneous biological data integration. In systematic tests on benchmark datasets and further experimental validation, UniBind effectively and scalably predicted the effects of SARS-CoV-2 spike protein variants on their binding affinities to the human ACE2 receptor, as well as to SARS-CoV-2 neutralizing monoclonal antibodies. Furthermore, in a cross-species analysis, UniBind could be applied to predict host susceptibility to SARS-CoV-2 variants and to predict future viral variant evolutionary trends. This in silico approach has the potential to serve as an early warning system for problematic emerging SARS-CoV-2 variants, as well as to facilitate research on protein-protein interactions in general.

Predicting the Effect of Single Mutations on Protein Stability and Binding with Respect to Types of Mutations

The development of methods and algorithms to predict the effect of mutations on protein stability, protein-protein interaction, and protein-DNA/RNA binding is necessitated by the needs of protein engineering and for understanding the molecular mechanism of disease-causing variants. The vast majority of the leading methods require a database of experimentally measured folding and binding free energy changes for training. These databases are collections of experimental data taken from scientific investigations typically aimed at probing the role of particular residues on the above-mentioned thermodynamic characteristics, i.e., the mutations are not introduced at random and do not necessarily represent mutations originating from single nucleotide variants (SNV). Thus, the reported performance of the leading algorithms assessed on these databases or other limited cases may not be applicable for predicting the effect of SNVs seen in the human population. Indeed, we demonstrate that the SNVs and non-SNVs are not equally presented in the corresponding databases, and the distribution of the free energy changes is not the same. It is shown that the Pearson correlation coefficients (PCCs) of folding and binding free energy changes obtained in cases involving SNVs are smaller than for non-SNVs, indicating that caution should be used in applying them to reveal the effect of human SNVs. Furthermore, it is demonstrated that some methods are sensitive to the chemical nature of the mutations, resulting in PCCs that differ by a factor of four across chemically different mutations. All methods are found to underestimate the energy changes by roughly a factor of 2.

EnCPdock: a web-interface for direct conjoint comparative analyses of complementarity and binding energetics in inter-protein associations

CONTEXT

Protein-protein interaction (PPI) is a key component linked to virtually all cellular processes. Be it an enzyme catalysis ('classic type functions' of proteins) or a signal transduction ('non-classic'), proteins generally function involving stable or quasi-stable multi-protein associations. The physical basis for such associations is inherent in the combined effect of shape and electrostatic complementarities (Sc, EC) of the interacting protein partners at their interface, which provides indirect probabilistic estimates of the stability and affinity of the interaction. While Sc is a necessary criterion for inter-protein associations, EC can be favorable as well as disfavored (e.g., in transient interactions). Estimating equilibrium thermodynamic parameters (∆G_binding, K_d) by experimental means is costly and time consuming, thereby opening windows for computational structural interventions. Attempts to empirically probe ∆G_binding from coarse-grain structural descriptors (primarily, surface area based terms) have lately been overtaken by physics-based, knowledge-based and their hybrid approaches (MM/PBSA, FoldX, etc.) that directly compute ∆G_binding without involving intermediate structural descriptors.

METHODS

Here, we present EnCPdock ( https://www.scinetmol.in/EnCPdock/ ), a user-friendly web-interface for the direct conjoint comparative analyses of complementarity and binding energetics in proteins. EnCPdock returns an AI-predicted ∆G_binding computed by combining complementarity (Sc, EC) and other high-level structural descriptors (input feature vectors), and renders a prediction accuracy comparable to the state-of-the-art. EnCPdock further locates a PPI complex in terms of its {Sc, EC} values (taken as an ordered pair) in the two-dimensional complementarity plot (CP). In addition, it also generates mobile molecular graphics of the interfacial atomic contact network for further analyses. EnCPdock also furnishes individual feature trends along with the relative probability estimates (Pr_fmax) of the obtained feature-scores with respect to the events of their highest observed frequencies. Together, these functionalities are of real practical use for structural tinkering and intervention as might be relevant in the design of targeted protein-interfaces. Combining all its features and applications, EnCPdock presents a unique online tool that should be beneficial to structural biologists and researchers across related fraternities.

Deep Local Analysis deconstructs protein-protein interfaces and accurately estimates binding affinity changes upon mutation

MOTIVATION

The spectacular recent advances in protein and protein complex structure prediction hold promise for reconstructing interactomes at large-scale and residue resolution. Beyond determining the 3D arrangement of interacting partners, modeling approaches should be able to unravel the impact of sequence variations on the strength of the association.

RESULTS

In this work, we report on Deep Local Analysis, a novel and efficient deep learning framework that relies on a strikingly simple deconstruction of protein interfaces into small locally oriented residue-centered cubes and on 3D convolutions recognizing patterns within cubes. Merely based on the two cubes associated with the wild-type and the mutant residues, DLA accurately estimates the binding affinity change for the associated complexes. It achieves a Pearson correlation coefficient of 0.735 on about 400 mutations on unseen complexes. Its generalization capability on blind datasets of complexes is higher than the state-of-the-art methods. We show that taking into account the evolutionary constraints on residues contributes to predictions. We also discuss the influence of conformational variability on performance. Beyond the predictive power on the effects of mutations, DLA is a general framework for transferring the knowledge gained from the available non-redundant set of complex protein structures to various tasks. For instance, given a single partially masked cube, it recovers the identity and physicochemical class of the central residue. Given an ensemble of cubes representing an interface, it predicts the function of the complex.

AVAILABILITY AND IMPLEMENTATION

Source code and models are available at http://gitlab.lcqb.upmc.fr/DLA/DLA.git.

Protein structure-based evaluation of missense variants: Resources, challenges and future directions

We provide an overview of the methods that can be used for protein structure-based evaluation of missense variants. The algorithms can be broadly divided into those that calculate the difference in free energy (ΔΔG) between the wild type and variant structures and those that use structural features to predict the damaging effect of a variant without providing a ΔΔG. A wide range of machine learning approaches have been employed to develop those algorithms. We also discuss challenges and opportunities for variant interpretation in view of the recent breakthrough in three-dimensional structural modelling using deep learning.

Recent advances in predicting and modeling protein-protein interactions

Protein-protein interactions (PPIs) drive biological processes, and disruption of PPIs can cause disease. With recent breakthroughs in structure prediction and a deluge of genomic sequence data, computational methods to predict PPIs and model spatial structures of protein complexes are now approaching the accuracy of experimental approaches for permanent interactions and show promise for elucidating transient interactions. As we describe here, the key to this success is rich evolutionary information deciphered from thousands of homologous sequences that coevolve in interacting partners. This covariation signal, revealed by sophisticated statistical and machine learning (ML) algorithms, predicts physiological interactions. Accurate artificial intelligence (AI)-based modeling of protein structures promises to provide accurate 3D models of PPIs at a proteome-wide scale.

Computational and artificial intelligence-based methods for antibody development

Due to their high target specificity and binding affinity, therapeutic antibodies are currently the largest class of biotherapeutics. The traditional largely empirical antibody development process is, while mature and robust, cumbersome and has significant limitations. Substantial recent advances in computational and artificial intelligence (AI) technologies are now starting to overcome many of these limitations and are increasingly integrated into development pipelines. Here, we provide an overview of AI methods relevant for antibody development, including databases, computational predictors of antibody properties and structure, and computational antibody design methods with an emphasis on machine learning (ML) models, and the design of complementarity-determining region (CDR) loops, antibody structural components critical for binding.

Predicting binding affinity changes from long-distance mutations using molecular dynamics simulations and Rosetta

Computationally modeling how mutations affect protein-protein binding not only helps uncover the biophysics of protein interfaces, but also enables the redesign and optimization of protein interactions. Traditional high-throughput methods for estimating binding free energy changes are currently limited to mutations directly at the interface due to difficulties in accurately modeling how long-distance mutations propagate their effects through the protein structure. However, the modeling and design of such mutations is of substantial interest as it allows for greater control and flexibility in protein design applications. We have developed a method that combines high-throughput Rosetta-based side-chain optimization with conformational sampling using classical molecular dynamics simulations, finding significant improvements in our ability to accurately predict long-distance mutational perturbations to protein binding. Our approach uses an analytical framework grounded in alchemical free energy calculations while enabling exploration of a vastly larger sequence space. When comparing to experimental data, we find that our method can predict internal long-distance mutational perturbations with a level of accuracy similar to that of traditional methods in predicting the effects of mutations at the protein-protein interface. This work represents a new and generalizable approach to optimize protein free energy landscapes for desired biological functions.

Assessment of Different Parameters on the Accuracy of Computational Alanine Scanning of Protein-Protein Complexes with the Molecular Mechanics/Generalized Born Surface Area Method

Computational alanine scanning with the molecular mechanics generalized Born surface area (MM/GBSA) method constitutes a widely used approach for identifying critical residues at protein-protein interfaces. Despite its popularity, the MM/GBSA method still has certain drawbacks due to its dependence on many factors. Here, we performed a systematical study on the impact of four different parameters, namely, the internal dielectric constant, the generalized Born model, the entropic term, and the inclusion of structural waters on the accuracy of computational alanine scanning calculations with the MM/GBSA method. Our results show that the internal dielectric constant is the most critical parameter for getting accurate predictions. The introduction of entropy and interfacial water molecules decreased the quality of the predictions, while the generalized Born model had little to no effect. Considering the significance of the internal dielectric value, we proposed a methodology based on the energetic predominance of a particular set of amino acids at the protein-protein interface for selecting an appropriate value for this variable. We hope that these results serve as a guideline for future studies of protein-protein complexes using the MM/GBSA method.

Measures of epitope binding degeneracy from T cell receptor repertoires

Adaptive immunity is driven by specific binding of hypervariable receptors to diverse molecular targets. The sequence diversity of receptors and targets are both individually known but because multiple receptors can recognize the same target, a measure of the effective "functional" diversity of the human immune system has remained elusive. Here, we show that sequence near-coincidences within T cell receptors that bind specific epitopes provide a new window into this problem and allow the quantification of how binding probability covaries with sequence. We find that near-coincidence statistics within epitope-specific repertoires imply a measure of binding degeneracy to amino acid changes in receptor sequence that is consistent across disparate experiments. Paired data on both chains of the heterodimeric receptor are particularly revealing since simultaneous near-coincidences are rare and we show how they can be exploited to estimate the number of epitope responses that created the memory compartment. In addition, we find that paired-chain coincidences are strongly suppressed across donors with different human leukocyte antigens, evidence for a central role of antigen-driven selection in making paired chain receptors public. These results demonstrate the power of coincidence analysis to reveal the sequence determinants of epitope binding in receptor repertoires.

Growing ecosystem of deep learning methods for modeling protein-protein interactions

Numerous cellular functions rely on protein-protein interactions. Efforts to comprehensively characterize them remain challenged however by the diversity of molecular recognition mechanisms employed within the proteome. Deep learning has emerged as a promising approach for tackling this problem by exploiting both experimental data and basic biophysical knowledge about protein interactions. Here, we review the growing ecosystem of deep learning methods for modeling protein interactions, highlighting the diversity of these biophysically informed models and their respective trade-offs. We discuss recent successes in using representation learning to capture complex features pertinent to predicting protein interactions and interaction sites, geometric deep learning to reason over protein structures and predict complex structures, and generative modeling to design de novo protein assemblies. We also outline some of the outstanding challenges and promising new directions. Opportunities abound to discover novel interactions, elucidate their physical mechanisms, and engineer binders to modulate their functions using deep learning and, ultimately, unravel how protein interactions orchestrate complex cellular behaviors.

Analysis of conformational stability of interacting residues in protein binding interfaces

After approximately 60 years of work, the protein folding problem has recently seen rapid advancement thanks to the inventions of AlphaFold and RoseTTAFold, which are machine-learning algorithms capable of reliably predicting protein structures from their sequences. A key component in their success was the inclusion of pairwise interaction information between residues. As research focus shifts towards developing algorithms to design and engineer binding proteins, it is likely that knowledge of interaction features at protein interfaces can improve predictions. Here, 574 protein complexes were analyzed to identify the stability features of their pairwise interactions, revealing that interactions between pre-stabilized residues are a selected feature in protein binding interfaces. In a retrospective analysis of 475 de novo designed binding proteins with an experimental success rate of 19%, inclusion of pairwise interaction pre-stabilization parameters increased the frequency of identifying experimentally successful binders to 40%.

Emerging Dominant SARS-CoV-2 Variants

Accurate and reliable forecasting of emerging dominant severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants enables policymakers and vaccine makers to get prepared for future waves of infections. The last three waves of SARS-CoV-2 infections caused by dominant variants, Omicron (BA.1), BA.2, and BA.4/BA.5, were accurately foretold by our artificial intelligence (AI) models built with biophysics, genotyping of viral genomes, experimental data, algebraic topology, and deep learning. On the basis of newly available experimental data, we analyzed the impacts of all possible viral spike (S) protein receptor-binding domain (RBD) mutations on the SARS-CoV-2 infectivity. Our analysis sheds light on viral evolutionary mechanisms, i.e., natural selection through infectivity strengthening and antibody resistance. We forecast that BP.1, BL*, BA.2.75*, BQ.1*, and particularly BN.1* have a high potential to become the new dominant variants to drive the next surge. Our key projection about these variants dominance made on Oct. 18, 2022 (see arXiv:2210.09485) became reality in late November 2022.

Predicting mutational function using machine learning

Genetic variations are one of the major causes of phenotypic variations between human individuals. Although beneficial as being the substrate of evolution, germline mutations may cause diseases, including Mendelian diseases and complex diseases such as diabetes and heart diseases. Mutations occurring in somatic cells are a main cause of cancer and likely cause age-related phenotypes and other age-related diseases. Because of the high abundance of genetic variations in the human genome, i.e., millions of germline variations per human subject and thousands of additional somatic mutations per cell, it is technically challenging to experimentally verify the function of every possible mutation and their interactions. Significant progress has been made to solve this problem using computational approaches, especially machine learning (ML). Here, we review the progress and achievements made in recent years in this field of research. We classify the computational models in two ways: one according to their prediction goals including protein structural alterations, gene expression changes, and disease risks, and the other according to their methodologies, including non-machine learning methods, classical machine learning methods, and deep neural network methods. For models in each category, we discuss their architecture, prediction accuracy, and potential limitations. This review provides new insights into the applications and future directions of computational approaches in understanding the role of mutations in aging and disease.

PROT-ON: A structure-based detection of designer PROTein interface MutatiONs

The mutation-induced changes across protein-protein interfaces have often been observed to lead to severe diseases. Therefore, several computational tools have been developed to predict the impact of such mutations. Among these tools, FoldX and EvoEF1 stand out as fast and accurate alternatives. Expanding on the capabilities of these tools, we have developed the PROT-ON (PROTein-protein interface mutatiONs) framework, which aims at delivering the most critical protein interface mutations that can be used to design new protein binders. To realize this aim, PROT-ON takes the 3D coordinates of a protein dimer as an input. Then, it probes all possible interface mutations on the selected protein partner with EvoEF1 or FoldX. The calculated mutational energy landscape is statistically analyzed to find the most enriching and depleting mutations. Afterward, these extreme mutations are filtered out according to stability and optionally according to evolutionary criteria. The final remaining mutation list is presented to the user as the designer mutation set. Together with this set, PROT-ON provides several residue- and energy-based plots, portraying the synthetic energy landscape of the probed mutations. The stand-alone version of PROT-ON is deposited at https://github.com/CSB-KaracaLab/prot-on. The users can also use PROT-ON through our user-friendly web service http://proton.tools.ibg.edu.tr:8001/ (runs with EvoEF1 only). Considering its speed and the range of analysis provided, we believe that PROT-ON presents a promising means to estimate designer mutations.

Case report: A gain-of-function of hamartin may lead to a distinct "inverse TSC1-hamartin" phenotype characterized by reduced cell growth

Mutations of TSC1 and TSC2 genes cause classical Tuberous Sclerosis Complex (TSC), a neurocutaneous disorder characterized by a tendency to develop hamartias, hamartomas, and other tumors. We herein report on a girl, now aged 5 years, who presented a previously unreported, distinct clinical phenotype consisting of primary microcephaly (head circumference = 40 cm, -5.6 standard deviations), brain anomalies including hypoplasia of the corpus callosum (with a residual draft of the genu), simplified parieto-temporal gyral pattern, colpocephaly with ectasia of the temporal ventricular horns, intellectual disability, and a general pattern of reduced growth (with weight and height < 3rd centiles). No classical features of TSC were recorded; the girl harbored a novel missense variant in TSC1 (c.611G > A). We hypothesize that her clinical phenotype could be related to a "gain-of-function" of the TSC1 protein product hamartin, causing an increase in the effects of the protein on inhibition of its intracellular targets (i.e., mTORC or RAC1 pathways), resulting in a distinct "inverse TSC1-hamartin" phenotype characterized by reduced growth of cells instead of the more classical predisposition to increased cell growth.

Amino acid substitutions in human growth hormone affect secondary structure and receptor binding

The interaction between human Growth Hormone (hGH) and hGH Receptor (hGHR) has basic relevance to cancer and growth disorders, and hGH is the scaffold for Pegvisomant, an anti-acromegaly therapeutic. For the latter reason, hGH has been extensively engineered by early workers to improve binding and other properties. We are particularly interested in E174 which belongs to the hGH zinc-binding triad; the substitution E174A is known to significantly increase binding, but to now no explanation has been offered. We generated this and several computationally-selected single-residue substitutions at the hGHR-binding site of hGH. We find that, while many successfully slow down dissociation of the hGH-hGHR complex once bound, they also slow down the association of hGH to hGHR. The E174A substitution induces a change in the Circular Dichroism spectrum that suggests the appearance of coiled-coiling. Here we show that E174A increases affinity of hGH against hGHR because the off-rate is slowed down more than the on-rate. For E174Y (and certain mutations at other sites) the slowdown in on-rate was greater than that of the off-rate, leading to decreased affinity. The results point to a link between structure, zinc binding, and hGHR-binding affinity in hGH.

Beyond sequence: Structure-based machine learning

Recent breakthroughs in protein structure prediction demarcate the start of a new era in structural bioinformatics. Combined with various advances in experimental structure determination and the uninterrupted pace at which new structures are published, this promises an age in which protein structure information is as prevalent and ubiquitous as sequence. Machine learning in protein bioinformatics has been dominated by sequence-based methods, but this is now changing to make use of the deluge of rich structural information as input. Machine learning methods making use of structures are scattered across literature and cover a number of different applications and scopes; while some try to address questions and tasks within a single protein family, others aim to capture characteristics across all available proteins. In this review, we look at the variety of structure-based machine learning approaches, how structures can be used as input, and typical applications of these approaches in protein biology. We also discuss current challenges and opportunities in this all-important and increasingly popular field.

Machine learning methods for protein-protein binding affinity prediction in protein design

Protein-protein interactions govern a wide range of biological activity. A proper estimation of the protein-protein binding affinity is vital to design proteins with high specificity and binding affinity toward a target protein, which has a variety of applications including antibody design in immunotherapy, enzyme engineering for reaction optimization, and construction of biosensors. However, experimental and theoretical modelling methods are time-consuming, hinder the exploration of the entire protein space, and deter the identification of optimal proteins that meet the requirements of practical applications. In recent years, the rapid development in machine learning methods for protein-protein binding affinity prediction has revealed the potential of a paradigm shift in protein design. Here, we review the prediction methods and associated datasets and discuss the requirements and construction methods of binding affinity prediction models for protein design.

Persistent Laplacian projected Omicron BA.4 and BA.5 to become new dominating variants

Due to its high transmissibility, Omicron BA.1 ousted the Delta variant to become a dominating variant in late 2021 and was replaced by more transmissible Omicron BA.2 in March 2022. An important question is which new variants will dominate in the future. Topology-based deep learning models have had tremendous success in forecasting emerging variants in the past. However, topology is insensitive to homotopic shape evolution in virus-human protein-protein binding, which is crucial to viral evolution and transmission. This challenge is tackled with persistent Laplacian, which is able to capture both the topological change and homotopic shape evolution of data. Persistent Laplacian-based deep learning models are developed to systematically evaluate variant infectivity. Our comparative analysis of Alpha, Beta, Gamma, Delta, Lambda, Mu, and Omicron BA.1, BA.1.1, BA.2, BA.2.11, BA.2.12.1, BA.3, BA.4, and BA.5 unveils that Omicron BA.2.11, BA.2.12.1, BA.3, BA.4, and BA.5 are more contagious than BA.2. In particular, BA.4 and BA.5 are about 36% more infectious than BA.2 and are projected to become new dominant variants by natural selection. Moreover, the proposed models outperform the state-of-the-art methods on three major benchmark datasets for mutation-induced protein-protein binding free energy changes. Our key projection about BA4 and BA.5's dominance made on May 1, 2022 (see arXiv:2205.00532) became a reality in late June 2022.

D3AI-Spike: A deep learning platform for predicting binding affinity between SARS-CoV-2 spike receptor binding domain with multiple amino acid mutations and human angiotensin-converting enzyme 2

The number of SARS-CoV-2 spike Receptor Binding Domain (RBD) with multiple amino acid mutations is huge due to random mutations and combinatorial explosions, making it almost impossible to experimentally determine their binding affinities to human angiotensin-converting enzyme 2 (hACE2). Although computational prediction is an alternative way, there is still no online platform to predict the mutation effect of RBD on the hACE2 binding affinity until now. In this study, we developed a free online platform based on deep learning models, namely D3AI-Spike, for quickly predicting binding affinity between spike RBD mutants and hACE2. The models based on CNN and CNN-RNN methods have the concordance index of around 0.8. Overall, the test results of the models are in agreement with the experimental data. To further evaluate the prediction power of D3AI-Spike, we predicted and experimentally determined the binding affinity of a VUM (variants under monitoring) variant IHU (B.1.640.2), which has fourteen amino acid substitutions, including N501Y and E484K, and 9 deletions located in the spike protein. The predicted average affinity score for wild-type RBD and IHU to hACE2 are 0.483 and 0.438, while the determined K_aff values are 5.39 ± 0.38 × 10⁷ L/mol and 1.02 ± 0.47 × 10⁷ L/mol, respectively, demonstrating the strong predictive power of D3AI-Spike. We think D3AI-Spike will be helpful to the viral transmission prediction for the new emerging SARS-CoV-2 variants. D3AI-Spike is now available free of charge at https://www.d3pharma.com/D3Targets-2019-nCoV/D3AI-Spike/index.php.

Densest subgraph-based methods for protein-protein interaction hot spot prediction

Background

Hot spots play an important role in protein binding analysis. The residue interaction network is a key point in hot spot prediction, and several graph theory-based methods have been proposed to detect hot spots. Although the existing methods can yield some interesting residues by network analysis, low recall has limited their abilities in finding more potential hot spots.

Result

In this study, we develop three graph theory-based methods to predict hot spots from only a single residue interaction network. We detect the important residues by finding subgraphs with high densities, i.e., high average degrees. Generally, a high degree implies a high binding possibility between protein chains, and thus a subgraph with high density usually relates to binding sites that have a high rate of hot spots. By evaluating the results on 67 complexes from the SKEMPI database, our methods clearly outperform existing graph theory-based methods on recall and F-score. In particular, our main method, Min-SDS, has an average recall of over 0.665 and an f2-score of over 0.364, while the recall and f2-score of the existing methods are less than 0.400 and 0.224, respectively.

Conclusion

The Min-SDS method performs best among all tested methods on the hot spot prediction problem, and all three of our methods provide useful approaches for analyzing bionetworks. In addition, the densest subgraph-based methods predict hot spots with only one residue interaction network, which is constructed from spatial atomic coordinate data to mitigate the shortage of data from wet-lab experiments.

Quantification and Neutralization of the Interfacial Electrostatic Potential and Visualization of the Dispersion Interaction in Visualization of the Interfacial Electrostatic Complementarity

Visualization of the interfacial electrostatic complementarity (VIINEC) is a quantum chemistry-based method to examine protein-protein interactions (PPI). In VIINEC, the electrostatic complementarity between proteins at the interface is visually and quantitatively evaluated using the partial electrostatic potential (pESP), which is defined based on the fragment molecular orbital method. In this work, new quantification and neutralization methods of the pESP were proposed together with a method to visualize the dispersion interaction. The reliability and efficiency of these methods were evaluated using 17 models of the complex. It was found that the quantification of the electrostatic complementarity with the pESP using the new neutralization method has a high correlation with the interaction energy, supporting the reliability of VIINEC. As an illustrative example, the PPI between a major histocompatibility complex class I molecule and a T-cell receptor was examined, which demonstrated the value of VIINEC in chemical and biological research.

Emerging dominant SARS-CoV-2 variants

Accurate and reliable forecasting of emerging dominant severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants enables policymakers and vaccine makers to get prepared for future waves of infections. The last three waves of SARS-CoV-2 infections caused by dominant variants Omicron (BA.1), BA.2, and BA.4/BA.5 were accurately foretold by our artificial intelligence (AI) models built with biophysics, genotyping of viral genomes, experimental data, algebraic topology, and deep learning. Based on newly available experimental data, we analyzed the impacts of all possible viral spike (S) protein receptor-binding domain (RBD) mutations on the SARS-CoV-2 infectivity. Our analysis sheds light on viral evolutionary mechanisms, i.e., natural selection through infectivity strengthening and antibody resistance. We forecast that BA.2.10.4, BA.2.75, BQ.1.1, and particularly, BA.2.75+R346T, have high potential to become new dominant variants to drive the next surge.