Local Interaction Signal Analysis Predicts Protein-Protein Binding Affinity

Assessing the functional impact of protein binding site definition

Many biomedical applications, such as classification of binding specificities or bioengineering, depend on the accurate definition of protein binding interfaces. Depending on the choice of method used, substantially different sets of residues can be classified as belonging to the interface of a protein. A typical approach used to verify these definitions is to mutate residues and measure the impact of these changes on binding. Besides the lack of exhaustive data, this approach also suffers from the fundamental problem that a mutation introduces an unknown amount of alteration into an interface, which potentially alters the binding characteristics of the interface. In this study we explore the impact of alternative binding site definitions on the ability of a protein to recognize its cognate ligand using a pharmacophore approach, which does not affect the interface. The study also shows that methods for protein binding interface predictions should perform above approximately F-score = 0.7 accuracy level to capture the biological function of a protein.

Software for Predicting Binding Free Energy of Protein-Protein Complexes and Their Mutants

Protein-protein binding affinity prediction is important for understanding complex biochemical pathways and to uncover protein interaction networks. Quantitative estimation of the binding affinity changes caused by mutations can provide critical information for protein function annotation and genetic disease diagnoses. The binding free energies of protein-protein complexes can be predicted using several computational tools. This chapter is a summary of software developed for the prediction of binding free energies for protein-protein complexes and their mutants.

A computational toolbox for the assembly yield of complex and heterogeneous structures

The self-assembly of complex structures from a set of non-identical building blocks is a hallmark of soft matter and biological systems, including protein complexes, colloidal clusters, and DNA-based assemblies. Predicting the dependence of the equilibrium assembly yield on the concentrations and interaction energies of building blocks is highly challenging, owing to the difficulty of computing the entropic contributions to the free energy of the many structures that compete with the ground state configuration. While these calculations yield well known results for spherically symmetric building blocks, they do not hold when the building blocks have internal rotational degrees of freedom. Here we present an approach for solving this problem that works with arbitrary building blocks, including proteins with known structure and complex colloidal building blocks. Our algorithm combines classical statistical mechanics with recently developed computational tools for automatic differentiation. Automatic differentiation allows efficient evaluation of equilibrium averages over configurations that would otherwise be intractable. We demonstrate the validity of our framework by comparison to molecular dynamics simulations of simple examples, and apply it to calculate the yield curves for known protein complexes and for the assembly of colloidal shells.

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 .

Cooperativity between H-bonds and tetrel bonds. Transformation of a noncovalent C⋯N tetrel bond to a covalent bond

The dimers and trimers formed by imidazole (IM) and F2TO (T = C, Si, Ge) are studied by ab initio calculations. IM can engage in either a NH⋯O H-bond with F2TO or a T⋯N tetrel bond (TB) with the π-hole above the T atom. The latter is a true noncovalent TB for T = C but is a much shorter and stronger covalent bond with F2SiO or F2GeO. When a second IM is added, the cooperativity emerging from its H-bond with the first IM makes it a stronger nucleophile, leading to two minima with F2CO. The first structure contains a long noncovalent C⋯O TB and there is a much shorter covalent bond in the other, with a small energy barrier separating them. The same sort of double minimum occurs when the two IM units are situated parallel to one another in a stacked geometry.

An updated dataset and a structure-based prediction model for protein-RNA binding affinity

Understanding the process of protein-RNA interaction is essential for structural biology. The thermodynamic process is an important part to uncover the protein-RNA interaction mechanism. The regulatory networks between protein and RNA in organisms are dominated by the binding or dissociation in the cells. Therefore, determining the binding affinity for protein-RNA complexes can help us to understand the regulation mechanism of protein-RNA interaction. Since it is time-consuming and labor-intensive to determine the binding affinity for protein-RNA complexes by experimental methods, it is necessary and urgent to develop computational methods to predict that. To develop a binding affinity prediction model, first we update the dataset of protein-RNA binding affinity benchmark (PRBAB), which includes 145 complexes now. Second, we extract the structural features based on complex structure, and then we analyze and select the representative structural features to train the regression model. Third, we random select the subset from the PRBAB2.0 to fit the protein-RNA binding affinity determined by experiment. In the end, we tested our model on the nonredundant PDBbind dataset, and the results showed that Pearson correlation coefficient r = .57 and RMSE = 2.51 kcal/mol. The Pearson correlation coefficient achieves 0.7 while removing 5 complex structures with modified residues/nucleotides and metal ions. While testing on ProNAB, the results showed that 71.60% of the prediction achieves Pearson correlation coefficient r = .61 and RMSE = 1.56 kcal/mol with experiment values.

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.

AREA-AFFINITY: A Web Server for Machine Learning-Based Prediction of Protein-Protein and Antibody-Protein Antigen Binding Affinities

Protein-Protein binding affinity reflects the binding strength between the binding partners. The prediction of protein-protein binding affinity is important for elucidating protein functions and also for designing protein-based therapeutics. The geometric characteristics such as area (both interface and surface areas) in the structure of a protein-protein complex play an important role in determining protein-protein interactions and their binding affinity. Here, we present a free web server for academic use, AREA-AFFINITY, for prediction of protein-protein or antibody-protein antigen binding affinity based on interface and surface areas in the structure of a protein-protein complex. AREA-AFFINITY implements 60 effective area-based protein-protein affinity predictive models and 37 effective area-based models specific for antibody-protein antigen binding affinity prediction developed in our recent studies. These models take into consideration the roles of interface and surface areas in binding affinity by using areas classified according to different amino acid types with different biophysical nature. The models with the best performances integrate machine learning methods such as neural network or random forest. These newly developed models have superior or comparable performance compared to the commonly used existing methods. AREA-AFFINITY is available for free at: https://affinity.cuhk.edu.cn/.

Extracting Quantitative Information at Quantum Mechanical Level from Noncovalent Interaction Index Analyses

The noncovalent interaction (NCI) index is nowadays a well-known strategy to detect NCIs in molecular systems. Even though it initially provided only qualitative descriptions, the technique has been recently extended to also extract quantitative information. To accomplish this task, integrals of powers of the electron distribution were considered, with the requirement that the overall electron density can be clearly decomposed as sum of distinct fragment contributions to enable the definition of the (noncovalent) integration region. So far, this was done by only exploiting approximate promolecular electron densities, which are given by the sum of spherically averaged atomic electron distributions and thus represent too crude approximations. Therefore, to obtain more quantum mechanically (QM) rigorous results from NCI index analyses, in this work, we propose to use electron densities obtained through the transfer of extremely localized molecular orbitals (ELMOs) or through the recently developed QM/ELMO embedding technique. Although still approximate, the electron distributions resulting from the abovementioned methods are fully QM and, above all, are again partitionable into subunit contributions, which makes them completely suitable for the NCI integral approach. Therefore, we benchmarked the integrals resulting from NCI index analyses (both those based on the promolecular densities and those based on ELMO electron distributions) against interaction energies computed at a high quantum chemical level (in particular, at the coupled cluster level). The performed test calculations have indicated that the NCI integrals based on ELMO electron densities outperform the promolecular ones. Furthermore, it was observed that the novel quantitative NCI-(QM/)ELMO approach can be also profitably exploited both to characterize and evaluate the strength of specific interactions between ligand subunits and protein residues in protein-ligand complexes and to follow the evolution of NCIs along trajectories of molecular dynamics simulations. Although further methodological improvements are still possible, the new quantitative ELMO-based technique could be already exploited in situations in which fast and reliable assessments of NCIs are crucial, such as in computational high-throughput screenings for drug discovery.

Molecular Bases of Protein Antigenicity and Determinants of Immunogenicity, Anergy, and Mitogenicity

BACKGROUND

The immune system is able to recognize substances that originate from inside or outside the body and are potentially harmful. Foreign substances that bind to immune system components exhibit antigenicity and are defined as antigens. The antigens exhibiting immunogenicity can induce innate or adaptive immune responses and give rise to humoral or cell-mediated immunity. The antigens exhibiting mitogenicity can cross-link cell membrane receptors on B and T lymphocytes leading to cell proliferation. All antigens vary greatly in physicochemical features such as biochemical nature, structural complexity, molecular size, foreignness, solubility, and so on.

OBJECTIVE

Thus, this review aims to describe the molecular bases of protein-antigenicity and those molecular bases that lead to an immune response, lymphocyte proliferation, or unresponsiveness.

CONCLUSION

The epitopes of an antigen are located in surface areas; they are about 880-3,300 Da in size. They are protein, carbohydrate, or lipid in nature. Soluble antigens are smaller than 1 nm and are endocytosed less efficiently than particulate antigens. The more the structural complexity of an antigen increases, the more the antigenicity increases due to the number and variety of epitopes. The smallest immunogens are about 4,000-10,000 Da in size. The more phylogenetically distant immunogens are from the immunogen-recipient, the more immunogenicity increases. Antigens that are immunogens can trigger an innate or adaptive immune response. The innate response is induced by antigens that are pathogen-associated molecular patterns. Exogenous antigens, T Dependent or T Independent, induce humoral immunogenicity. TD protein-antigens require two epitopes, one sequential and one conformational to induce antibodies, whereas, TI non-protein-antigens require only one conformational epitope to induce low-affinity antibodies. Endogenous protein antigens require only one sequential epitope to induce cell-mediated immunogenicity.

Cold spots are universal in protein-protein interactions

Proteins interact with each other through binding interfaces that differ greatly in size and physico-chemical properties. Within the binding interface, a few residues called hot spots contribute the majority of the binding free energy and are hence irreplaceable. In contrast, cold spots are occupied by suboptimal amino acids, providing possibility for affinity enhancement through mutations. In this study, we identify cold spots due to cavities and unfavorable charge interactions in multiple protein-protein interactions (PPIs). For our cold spot analysis, we first use a small affinity database of PPIs with known structures and affinities and then expand our search to nearly 4000 homo- and heterodimers in the Protein Data Bank (PDB). We observe that cold spots due to cavities are present in nearly all PPIs unrelated to their binding affinity, while unfavorable charge interactions are relatively rare. We also find that most cold spots are located in the periphery of the binding interface, with high-affinity complexes showing fewer centrally located colds spots than low-affinity complexes. A larger number of cold spots is also found in non-cognate interactions compared to their cognate counterparts. Furthermore, our analysis reveals that cold spots are more frequent in homo-dimeric complexes compared to hetero-complexes, likely due to symmetry constraints imposed on sequences of homodimers. Finally, we find that glycines, glutamates, and arginines are the most frequent amino acids appearing at cold spot positions. Our analysis emphasizes the importance of cold spot positions to protein evolution and facilitates protein engineering studies directed at enhancing binding affinity and specificity in a wide range of applications.

Deep Local Analysis evaluates protein docking conformations with locally oriented cubes

MOTIVATION

With the recent advances in protein 3D structure prediction, protein interactions are becoming more central than ever before. Here, we address the problem of determining how proteins interact with one another. More specifically, we investigate the possibility of discriminating near-native protein complex conformations from incorrect ones by exploiting local environments around interfacial residues.

RESULTS

Deep Local Analysis (DLA)-Ranker is a deep learning framework applying 3D convolutions to a set of locally oriented cubes representing the protein interface. It explicitly considers the local geometry of the interfacial residues along with their neighboring atoms and the regions of the interface with different solvent accessibility. We assessed its performance on three docking benchmarks made of half a million acceptable and incorrect conformations. We show that DLA-Ranker successfully identifies near-native conformations from ensembles generated by molecular docking. It surpasses or competes with other deep learning-based scoring functions. We also showcase its usefulness to discover alternative interfaces.

AVAILABILITY AND IMPLEMENTATION

http://gitlab.lcqb.upmc.fr/dla-ranker/DLA-Ranker.git.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Structural insights into SARS-CoV-2 spike protein and its natural mutants found in Mexican population

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a newly emerged coronavirus responsible for coronavirus disease 2019 (COVID-19); it become a pandemic since March 2020. To date, there have been described three lineages of SARS-CoV-2 circulating worldwide, two of them are found among Mexican population, within these, we observed three mutations of spike (S) protein located at amino acids H49Y, D614G, and T573I. To understand if these mutations could affect the structural behavior of S protein of SARS-CoV-2, as well as the binding with S protein inhibitors (cepharanthine, nelfinavir, and hydroxychloroquine), molecular dynamic simulations and molecular docking were employed. It was found that these punctual mutations affect considerably the structural behavior of the S protein compared to wild type, which also affect the binding of its inhibitors into their respective binding site. Thus, further experimental studies are needed to explore if these affectations have an impact on drug-S protein binding and its possible clinical effect.

An expanded benchmark for antibody-antigen docking and affinity prediction reveals insights into antibody recognition determinants

Accurate predictive modeling of antibody-antigen complex structures and structure-based antibody design remain major challenges in computational biology, with implications for biotherapeutics, immunity, and vaccines. Through a systematic search for high-resolution structures of antibody-antigen complexes and unbound antibody and antigen structures, in conjunction with identification of experimentally determined binding affinities, we have assembled a non-redundant set of test cases for antibody-antigen docking and affinity prediction. This benchmark more than doubles the number of antibody-antigen complexes and corresponding affinities available in our previous benchmarks, providing an unprecedented view of the determinants of antibody recognition and insights into molecular flexibility. Initial assessments of docking and affinity prediction tools highlight the challenges posed by this diverse set of cases, which includes camelid nanobodies, therapeutic monoclonal antibodies, and broadly neutralizing antibodies targeting viral glycoproteins. This dataset will enable development of advanced predictive modeling and design methods for this therapeutically relevant class of protein-protein interactions.

Challenges in protein docking

Current developments in protein docking aim at improvement of applicability, accuracy and utility of modeling macromolecular complexes. The challenges include the need for greater emphasis on protein docking to molecules of different types, proper accounting for conformational flexibility upon binding, new promising methodologies based on residue co-evolution and deep learning, affinity prediction, and further development of fully automated docking servers. Importantly, new developments increasingly focus on realistic modeling of protein interactions in vivo, including crowded environment inside a cell, which involves multiple transient encounters, and propagating the system in time. This opinion paper offers the author's perspective on these challenges in structural modeling of protein interactions and the future of protein docking.

NCIPLOT4: Fast, Robust, and Quantitative Analysis of Noncovalent Interactions

The NonCovalent Interaction index (NCI) enables identification of attractive and repulsive noncovalent interactions from promolecular densities in a fast manner. However, the approach remained up to now qualitative, only providing visual information. We present a new version of NCIPLOT, NCIPLOT4, which allows quantifying the properties of the NCI regions (volume, charge) in small and big systems in a fast manner. Examples are provided of how this new twist enables characterization and retrieval of local information in supramolecular chemistry and biosystems at the static and dynamic levels.

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.

Development of R7BP inhibitors through cross-linking coupled mass spectrometry and integrated modeling

Protein-protein interaction (PPI) networks are known to be valuable targets for therapeutic intervention; yet the development of PPI modulators as next-generation drugs to target specific vertices, edges, and hubs has been impeded by the lack of structural information of many of the proteins and complexes involved. Building on recent advancements in cross-linking mass spectrometry (XL-MS), we describe an effective approach to obtain relevant structural data on R7BP, a master regulator of itch sensation, and its interfaces with other proteins in its network. This approach integrates XL-MS with a variety of modeling techniques to successfully develop antibody inhibitors of the R7BP and RGS7/Gβ5 duplex interaction. Binding and inhibitory efficiency are studied by surface plasmon resonance spectroscopy and through an R7BP-derived dominant negative construct. This approach may have broader applications as a tool to facilitate the development of PPI modulators in the absence of crystal structures or when structural information is limited.

FRETting about the affinity of bimolecular protein-protein interactions

Fluorescence resonance energy transfer (FRET) is a powerful tool to study macromolecular interactions such as protein-protein interactions (PPIs). Fluorescent protein (FP) fusions enable FRET-based PPI analysis of signaling pathways and molecular structure in living cells. Despite FRET's importance in PPI studies, FRET has seen limited use in quantifying the affinities of PPIs in living cells. Here, we have explored the relationship between FRET efficiency and PPI affinity over a wide range when expressed from a single plasmid system in Escherichia coli. Using live-cell microscopy and a set of 20 pairs of small interacting proteins, belonging to different structural folds and interaction affinities, we demonstrate that FRET efficiency can reliably measure the dissociation constant (K_D ) over a range of mM to nM. A 10-fold increase in the interaction affinity results in 0.05 unit increase in FRET efficiency, providing sufficient resolution to quantify large affinity differences (> 10-fold) using live-cell FRET. This approach provides a rapid and simple strategy for assessment of PPI affinities over a wide range and will have utility for high-throughput analysis of protein interactions.