Protein-protein docking benchmark version 4.0

SurfPro-NN: A 3D point cloud neural network for the scoring of protein-protein docking models based on surfaces features and protein language models

Protein-protein interactions (PPI) play a crucial role in numerous key biological processes, and the structure of protein complexes provides valuable clues for in-depth exploration of molecular-level biological processes. Protein-protein docking technology is widely used to simulate the spatial structure of proteins. However, there are still challenges in selecting candidate decoys that closely resemble the native structure from protein-protein docking simulations. In this study, we introduce a docking evaluation method based on three-dimensional point cloud neural networks named SurfPro-NN, which represents protein structures as point clouds and learns interaction information from protein interfaces by applying a point cloud neural network. With the continuous advancement of deep learning in the field of biology, a series of knowledge-rich pre-trained models have emerged. We incorporate protein surface representation models and language models into our approach, greatly enhancing feature representation capabilities and achieving superior performance in protein docking model scoring tasks. Through comprehensive testing on public datasets, we find that our method outperforms state-of-the-art deep learning approaches in protein-protein docking model scoring. Not only does it significantly improve performance, but it also greatly accelerates training speed. This study demonstrates the potential of our approach in addressing protein interaction assessment problems, providing strong support for future research and applications in the field of biology.

Encoding the space of protein-protein binding interfaces by artificial intelligence

The physical interactions between proteins are largely determined by the structural properties at their binding interfaces. It was found that the binding interfaces in distinctive protein complexes are highly similar. The structural properties underlying different binding interfaces could be further captured by artificial intelligence. In order to test this hypothesis, we broke protein-protein binding interfaces into pairs of interacting fragments. We employed a generative model to encode these interface fragment pairs in a low-dimensional latent space. After training, new conformations of interface fragment pairs were generated. We found that, by only using a small number of interface fragment pairs that were generated by artificial intelligence, we were able to guide the assembly of protein complexes into their native conformations. These results demonstrate that the conformational space of fragment pairs at protein-protein binding interfaces is highly degenerate. Features in this degenerate space can be well characterized by artificial intelligence. In summary, our machine learning method will be potentially useful to search for and predict the conformations of unknown protein-protein interactions.

A Survey of Deep Learning Methods for Estimating the Accuracy of Protein Quaternary Structure Models

The quality prediction of quaternary structure models of a protein complex, in the absence of its true structure, is known as the Estimation of Model Accuracy (EMA). EMA is useful for ranking predicted protein complex structures and using them appropriately in biomedical research, such as protein-protein interaction studies, protein design, and drug discovery. With the advent of more accurate protein complex (multimer) prediction tools, such as AlphaFold2-Multimer and ESMFold, the estimation of the accuracy of protein complex structures has attracted increasing attention. Many deep learning methods have been developed to tackle this problem; however, there is a noticeable absence of a comprehensive overview of these methods to facilitate future development. Addressing this gap, we present a review of deep learning EMA methods for protein complex structures developed in the past several years, analyzing their methodologies, data and feature construction. We also provide a prospective summary of some potential new developments for further improving the accuracy of the EMA methods.

ProInterVal: Validation of Protein-Protein Interfaces through Learned Interface Representations

Proteins are vital components of the biological world and serve a multitude of functions. They interact with other molecules through their interfaces and participate in crucial cellular processes. Disruption of these interactions can have negative effects on organisms, highlighting the importance of studying protein-protein interfaces for developing targeted therapies for diseases. Therefore, the development of a reliable method for investigating protein-protein interactions is of paramount importance. In this work, we present an approach for validating protein-protein interfaces using learned interface representations. The approach involves using a graph-based contrastive autoencoder architecture and a transformer to learn representations of protein-protein interaction interfaces from unlabeled data and then validating them through learned representations with a graph neural network. Our method achieves an accuracy of 0.91 for the test set, outperforming existing GNN-based methods. We demonstrate the effectiveness of our approach on a benchmark data set and show that it provides a promising solution for validating protein-protein interfaces.

ABAG-docking benchmark: a non-redundant structure benchmark dataset for antibody-antigen computational docking

Accurate prediction of antibody-antigen complex structures is pivotal in drug discovery, vaccine design and disease treatment and can facilitate the development of more effective therapies and diagnostics. In this work, we first review the antibody-antigen docking (ABAG-docking) datasets. Then, we present the creation and characterization of a comprehensive benchmark dataset of antibody-antigen complexes. We categorize the dataset based on docking difficulty, interface properties and structural characteristics, to provide a diverse set of cases for rigorous evaluation. Compared with Docking Benchmark 5.5, we have added 112 cases, including 14 single-domain antibody (sdAb) cases and 98 monoclonal antibody (mAb) cases, and also increased the proportion of Difficult cases. Our dataset contains diverse cases, including human/humanized antibodies, sdAbs, rodent antibodies and other types, opening the door to better algorithm development. Furthermore, we provide details on the process of building the benchmark dataset and introduce a pipeline for periodic updates to keep it up to date. We also utilize multiple complex prediction methods including ZDOCK, ClusPro, HDOCK and AlphaFold-Multimer for testing and analyzing this dataset. This benchmark serves as a valuable resource for evaluating and advancing docking computational methods in the analysis of antibody-antigen interaction, enabling researchers to develop more accurate and effective tools for predicting and designing antibody-antigen complexes. The non-redundant ABAG-docking structure benchmark dataset is available at https://github.com/Zhaonan99/Antibody-antigen-complex-structure-benchmark-dataset.

Discriminating physiological from non-physiological interfaces in structures of protein complexes: A community-wide study

Reliably scoring and ranking candidate models of protein complexes and assigning their oligomeric state from the structure of the crystal lattice represent outstanding challenges. A community-wide effort was launched to tackle these challenges. The latest resources on protein complexes and interfaces were exploited to derive a benchmark dataset consisting of 1677 homodimer protein crystal structures, including a balanced mix of physiological and non-physiological complexes. The non-physiological complexes in the benchmark were selected to bury a similar or larger interface area than their physiological counterparts, making it more difficult for scoring functions to differentiate between them. Next, 252 functions for scoring protein-protein interfaces previously developed by 13 groups were collected and evaluated for their ability to discriminate between physiological and non-physiological complexes. A simple consensus score generated using the best performing score of each of the 13 groups, and a cross-validated Random Forest (RF) classifier were created. Both approaches showed excellent performance, with an area under the Receiver Operating Characteristic (ROC) curve of 0.93 and 0.94, respectively, outperforming individual scores developed by different groups. Additionally, AlphaFold2 engines recalled the physiological dimers with significantly higher accuracy than the non-physiological set, lending support to the reliability of our benchmark dataset annotations. Optimizing the combined power of interface scoring functions and evaluating it on challenging benchmark datasets appears to be a promising strategy.

PointDE: Protein Docking Evaluation Using 3D Point Cloud Neural Network

Protein-protein interactions (PPIs) play essential roles in many vital movements and the determination of protein complex structure is helpful to discover the mechanism of PPI. Protein-protein docking is being developed to model the structure of the protein. However, there is still a challenge to selecting the near-native decoys generated by protein-protein docking. Here, we propose a docking evaluation method using 3D point cloud neural network named PointDE. PointDE transforms protein structure to the point cloud. Using the state-of-the-art point cloud network architecture and a novel grouping mechanism, PointDE can capture the geometries of the point cloud and learn the interaction information from the protein interface. On public datasets, PointDE surpasses the state-of-the-art method using deep learning. To further explore the ability of our method in different types of protein structures, we developed a new dataset generated by high-quality antibody-antigen complexes. The result in this antibody-antigen dataset shows the strong performance of PointDE, which will be helpful for the understanding of PPI mechanisms.

Using machine learning to predict protein-protein interactions between a zombie ant fungus and its carpenter ant host

Parasitic fungi produce proteins that modulate virulence, alter host physiology, and trigger host responses. These proteins, classified as a type of "effector," often act via protein-protein interactions (PPIs). The fungal parasite Ophiocordyceps camponoti-floridani (zombie ant fungus) manipulates Camponotus floridanus (carpenter ant) behavior to promote transmission. The most striking aspect of this behavioral change is a summit disease phenotype where infected hosts ascend and attach to an elevated position. Plausibly, interspecific PPIs drive aspects of Ophiocordyceps infection and host manipulation. Machine learning PPI predictions offer high-throughput methods to produce mechanistic hypotheses on how this behavioral manipulation occurs. Using D-SCRIPT to predict host-parasite PPIs, we found ca. 6000 interactions involving 2083 host proteins and 129 parasite proteins, which are encoded by genes upregulated during manipulated behavior. We identified multiple overrepresentations of functional annotations among these proteins. The strongest signals in the host highlighted neuromodulatory G-protein coupled receptors and oxidation-reduction processes. We also detected Camponotus structural and gene-regulatory proteins. In the parasite, we found enrichment of Ophiocordyceps proteases and frequent involvement of novel small secreted proteins with unknown functions. From these results, we provide new hypotheses on potential parasite effectors and host targets underlying zombie ant behavioral manipulation.

Spatial Layouts of Low-Entropy Hydration Shells Guide Protein Binding

Protein-protein binding enables orderly biological self-organization and is therefore considered a miracle of nature. Protein‒protein binding is driven by electrostatic forces, hydrogen bonding, van der Waals force, and hydrophobic interactions. Among these physical forces, only hydrophobic interactions can be considered long-range intermolecular attractions between proteins due to the electrostatic shielding of surrounding water molecules. Low-entropy hydration shells around proteins drive hydrophobic attraction among them that essentially coordinate protein‒protein binding. Here, an innovative method is developed for identifying low-entropy regions of hydration shells of proteins by screening off pseudohydrophilic groups on protein surfaces and revealing that large low-entropy regions of the hydration shells typically cover the binding sites of individual proteins. According to an analysis of determined protein complex structures, shape matching between a large low-entropy hydration shell region of a protein and that of its partner at the binding sites is revealed as a universal law. Protein‒protein binding is thus found to be mainly guided by hydrophobic collapse between the shape-matched low-entropy hydration shells that is verified by bioinformatics analyses of hundreds of structures of protein complexes, which cover four test systems. A simple algorithm is proposed to accurately predict protein binding sites.

GDockScore: a graph-based protein-protein docking scoring function

Summary

Protein complexes play vital roles in a variety of biological processes, such as mediating biochemical reactions, the immune response and cell signalling, with 3D structure specifying function. Computational docking methods provide a means to determine the interface between two complexed polypeptide chains without using time-consuming experimental techniques. The docking process requires the optimal solution to be selected with a scoring function. Here, we propose a novel graph-based deep learning model that utilizes mathematical graph representations of proteins to learn a scoring function (GDockScore). GDockScore was pre-trained on docking outputs generated with the Protein Data Bank biounits and the RosettaDock protocol, and then fine-tuned on HADDOCK decoys generated on the ZDOCK Protein Docking Benchmark. GDockScore performs similarly to the Rosetta scoring function on docking decoys generated using the RosettaDock protocol. Furthermore, state-of-the-art is achieved on the CAPRI score set, a challenging dataset for developing docking scoring functions.

Availability and implementation

The model implementation is available at https://gitlab.com/mcfeemat/gdockscore.

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

PIQLE: protein-protein interface quality estimation by deep graph learning of multimeric interaction geometries

Motivation

Accurate modeling of protein-protein interaction interface is essential for high-quality protein complex structure prediction. Existing approaches for estimating the quality of a predicted protein complex structural model utilize only the physicochemical properties or energetic contributions of the interacting atoms, ignoring evolutionarily information or inter-atomic multimeric geometries, including interaction distance and orientations.

Results

Here, we present PIQLE, a deep graph learning method for protein-protein interface quality estimation. PIQLE leverages multimeric interaction geometries and evolutionarily information along with sequence- and structure-derived features to estimate the quality of individual interactions between the interfacial residues using a multi-head graph attention network and then probabilistically combines the estimated quality for scoring the overall interface. Experimental results show that PIQLE consistently outperforms existing state-of-the-art methods including DProQA, TRScore, GNN-DOVE and DOVE on multiple independent test datasets across a wide range of evaluation metrics. Our ablation study and comparison with the self-assessment module of AlphaFold-Multimer repurposed for protein complex scoring reveal that the performance gains are connected to the effectiveness of the multi-head graph attention network in leveraging multimeric interaction geometries and evolutionary information along with other sequence- and structure-derived features adopted in PIQLE.

Availability and implementation

An open-source software implementation of PIQLE is freely available at https://github.com/Bhattacharya-Lab/PIQLE.

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Protein-protein interfaces in molecular glue-induced ternary complexes: classification, characterization, and prediction

Molecular glues are a class of small molecules that stabilize the interactions between proteins. Naturally occurring molecular glues are present in many areas of biology where they serve as central regulators of signaling pathways. Importantly, several clinical compounds act as molecular glue degraders that stabilize interactions between E3 ubiquitin ligases and target proteins, leading to their degradation. Molecular glues hold promise as a new generation of therapeutic agents, including those molecular glue degraders that can redirect the protein degradation machinery in a precise way. However, rational discovery of molecular glues is difficult in part due to the lack of understanding of the protein-protein interactions they stabilize. In this review, we summarize the structures of known molecular glue-induced ternary complexes and the interface properties. Detailed analysis shows different mechanisms of ternary structure formation. Additionally, we also review computational approaches for predicting protein-protein interfaces and highlight the promises and challenges. This information will ultimately help inform future approaches for rational molecular glue discovery.

Ebola virus disease: In vivo protection provided by the PAMP restricted TLR3 agonist rintatolimod and its mechanism of action

Ebola virus (EBOV) is a highly infectious and lethal pathogen responsible for sporadic self-limiting clusters of Ebola virus disease (EVD) in Central Africa capable of reaching epidemic status. 100% protection from lethal EBOV-Zaire in Balb/c mice was achieved by rintatolimod (Ampligen) at the well tolerated human clinical dose of 6 mg/kg. The data indicate that the mechanism of action is rintatolimod's dual ability to act as both a competitive decoy for the IID domain of VP35 blocking viral dsRNA sequestration and as a pathogen-associated molecular pattern (PAMP) restricted agonist for direct TLR3 activation but lacking RIG-1-like cytosolic helicase agonist properties. These data show promise for rintatolimod as a prophylactic therapy against human Ebola outbreaks.

PIQLE: protein-protein interface quality estimation by deep graph learning of multimeric interaction geometries

Accurate modeling of protein-protein interaction interface is essential for high-quality protein complex structure prediction. Existing approaches for estimating the quality of a predicted protein complex structural model utilize only the physicochemical properties or energetic contributions of the interacting atoms, ignoring evolutionarily information or inter-atomic multimeric geometries, including interaction distance and orientations. Here we present PIQLE, a deep graph learning method for protein-protein interface quality estimation. PIQLE leverages multimeric interaction geometries and evolutionarily information along with sequence- and structure-derived features to estimate the quality of the individual interactions between the interfacial residues using a multihead graph attention network and then probabilistically combines the estimated quality of the interfacial residues for scoring the overall interface. Experimental results show that PIQLE consistently outperforms existing state-of-the-art methods on multiple independent test datasets across a wide range of evaluation metrics. Our ablation study reveals that the performance gains are connected to the effectiveness of the multihead graph attention network in leveraging multimeric interaction geometries and evolutionary information along with other sequence- and structure-derived features adopted in PIQLE. An open-source software implementation of PIQLE, licensed under the GNU General Public License v3, is freely available at https://github.com/Bhattacharya-Lab/PIQLE .

MetaScore: A Novel Machine-Learning-Based Approach to Improve Traditional Scoring Functions for Scoring Protein-Protein Docking Conformations

Protein-protein interactions play a ubiquitous role in biological function. Knowledge of the three-dimensional (3D) structures of the complexes they form is essential for understanding the structural basis of those interactions and how they orchestrate key cellular processes. Computational docking has become an indispensable alternative to the expensive and time-consuming experimental approaches for determining the 3D structures of protein complexes. Despite recent progress, identifying near-native models from a large set of conformations sampled by docking-the so-called scoring problem-still has considerable room for improvement. We present MetaScore, a new machine-learning-based approach to improve the scoring of docked conformations. MetaScore utilizes a random forest (RF) classifier trained to distinguish near-native from non-native conformations using their protein-protein interfacial features. The features include physicochemical properties, energy terms, interaction-propensity-based features, geometric properties, interface topology features, evolutionary conservation, and also scores produced by traditional scoring functions (SFs). MetaScore scores docked conformations by simply averaging the score produced by the RF classifier with that produced by any traditional SF. We demonstrate that (i) MetaScore consistently outperforms each of the nine traditional SFs included in this work in terms of success rate and hit rate evaluated over conformations ranked among the top 10; (ii) an ensemble method, MetaScore-Ensemble, that combines 10 variants of MetaScore obtained by combining the RF score with each of the traditional SFs outperforms each of the MetaScore variants. We conclude that the performance of traditional SFs can be improved upon by using machine learning to judiciously leverage protein-protein interfacial features and by using ensemble methods to combine multiple scoring functions.

Increased slow dynamics defines ligandability of BTB domains

Efficient determination of protein ligandability, or the propensity to bind small-molecules, would greatly facilitate drug development for novel targets. Ligandability is currently assessed using computational methods that typically consider the static structural properties of putative binding sites or by experimental fragment screening. Here, we evaluate ligandability of conserved BTB domains from the cancer-relevant proteins LRF, KAISO, and MIZ1. Using fragment screening, we discover that MIZ1 binds multiple ligands. However, no ligands are uncovered for the structurally related KAISO or LRF. To understand the principles governing ligand-binding by BTB domains, we perform comprehensive NMR-based dynamics studies and find that only the MIZ1 BTB domain exhibits backbone µs-ms time scale motions. Interestingly, residues with elevated dynamics correspond to the binding site of fragment hits and recently defined HUWE1 interaction site. Our data argue that examining protein dynamics using NMR can contribute to identification of cryptic binding sites, and may support prediction of the ligandability of novel challenging targets.

A dynamical view of protein-protein complexes: Studies by molecular dynamics simulations

Protein-protein interactions are at the basis of many protein functions, and the knowledge of 3D structures of protein-protein complexes provides structural, mechanical and dynamical pieces of information essential to understand these functions. Protein-protein interfaces can be seen as stable, organized regions where residues from different partners form non-covalent interactions that are responsible for interaction specificity and strength. They are commonly described as a peripheral region, whose role is to protect the core region that concentrates the most contributing interactions, from the solvent. To get insights into the dynamics of protein-protein complexes, we carried out all-atom molecular dynamics simulations in explicit solvent on eight different protein-protein complexes of different functional class and interface size by taking into account the bound and unbound forms. On the one hand, we characterized structural changes upon binding of the proteins, and on the other hand we extensively analyzed the interfaces and the structural waters involved in the binding. Based on our analysis, in 6 cases out of 8, the interfaces rearranged during the simulation time, in stable and long-lived substates with alternative residue-residue contacts. These rearrangements are not restricted to side-chain fluctuations in the periphery but also affect the core interface. Finally, the analysis of the waters at the interface and involved in the binding pointed out the importance to take into account their role in the estimation of the interaction strength.

Targeting Undruggable Transcription Factors with PROTACs: Advances and Perspectives

Dysregulation of transcription factors has been implicated in a variety of human diseases. However, these proteins have traditionally been regarded as undruggable and only a handful of them have been successfully targeted by conventional small molecules. Moreover, the development of intrinsic and acquired resistance has hampered the clinical use of these agents. Over the past years, proteolysis-targeting chimeras (PROTACs) have shown great promise because of their potential for overcoming drug resistance and their ability to target previously undruggable proteins. Indeed, several small molecule-based PROTACs have demonstrated superior efficacy in therapy-resistant metastatic cancers. Nevertheless, it remains challenging to identify ligands for the majority of transcription factors. Given that transcription factors recognize short DNA motifs in a sequence-specific manner, multiple novel approaches exploit DNA motifs as warheads in PROTAC design for the degradation of aberrant transcription factors. These PROTACs pave the way for targeting undruggable transcription factors with potential therapeutic benefits.

TRScore: a three-dimensional RepVGG-based scoring method for ranking protein docking models

MOTIVATION

Protein-protein interactions (PPI) play important roles in cellular activities. Due to the technical difficulty and high cost of experimental methods, there are considerable interests towards the development of computational approaches, such as protein docking, to decipher PPI patterns. One of the important and difficult aspects in protein docking is recognizing near-native conformations from a set of decoys, but unfortunately traditional scoring functions still suffer from limited accuracy. Therefore, new scoring methods are pressingly needed in methodological and/or practical implications.

RESULTS

We present a new deep learning-based scoring method for ranking protein-protein docking models based on a three-dimensional (3D) RepVGG network, named TRScore. To recognize near-native conformations from a set of decoys, TRScore voxelizes the protein-protein interface into a 3D grid labeled by the number of atoms in different physicochemical classes. Benefiting from the deep convolutional RepVGG architecture, TRScore can effectively capture the subtle differences between energetically favorable near-native models and unfavorable non-native decoys without needing extra information. TRScore was extensively evaluated on diverse test sets including protein-protein docking benchmark 5.0 update set, DockGround decoy set, as well as realistic CAPRI decoy set, and overall obtained a significant improvement over existing methods in cross validation and independent evaluations.

AVAILABILITY

Codes available at: https://github.com/BioinformaticsCSU/TRScore.

Recent advances in predicting protein-protein interactions with the aid of artificial intelligence algorithms

Protein-protein interactions (PPIs) are essential in the regulation of biological functions and cell events, therefore understanding PPIs have become a key issue to understanding the molecular mechanism and investigating the design of drugs. Here we highlight the major developments in computational methods developed for predicting PPIs by using types of artificial intelligence algorithms. The first part introduces the source of experimental PPI data. The second part is devoted to the PPI prediction methods based on sequential information. The third part covers representative methods using structural information as the input feature. The last part is methods designed by combining different types of features. For each part, the state-of-the-art computational PPI prediction methods are reviewed in an inclusive view. Finally, we discuss the flaws existing in this area and future directions of next-generation algorithms.

CSM-AB: graph-based antibody-antigen binding affinity prediction and docking scoring function

MOTIVATION

Understanding antibody-antigen interactions is key to improving their binding affinities and specificities. While experimental approaches are fundamental for developing new therapeutics, computational methods can provide quick assessment of binding landscapes, guiding experimental design. Despite this, little effort has been devoted to accurately predicting the binding affinity between antibodies and antigens and to develop tailored docking scoring functions for this type of interaction. Here, we developed CSM-AB, a machine learning method capable of predicting antibody-antigen binding affinity by modelling interaction interfaces as graph-based signatures.

RESULTS

CSM-AB outperformed alternative methods achieving a Pearson's correlation of up to 0.64 on blind tests. We also show CSM-AB can accurately rank near-native poses, working effectively as a docking scoring function. We believe CSM-AB will be an invaluable tool to assist in the development of new immunotherapies.

AVAILABILITY AND IMPLEMENTATION

CSM-AB is freely available as a user-friendly web interface and API at http://biosig.unimelb.edu.au/csm_ab/datasets.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

PEPscan: A Broad Spectrum Approach for the Characterization of Protein-Binder Interactions?

In a previous study, we have shown that PEPscan can provide a cheap and rapid means to identify candidate interfering peptides (IPs), i.e., peptides able to disrupt a target protein-protein interaction. PEPscan was shown to be effective in identifying a limited number of candidate IPs specific to the target interaction. Here, we investigate the results of 14 new PEPscan experiments for protein complexes of known 3D structures. We show that for almost all complexes, PEPscan is able to identify candidate IPs that are located at the protein-protein interface. The information it provides about the binding site seems, however, too ambiguous to be exploited in a simple manner to assist the modeling of protein complexes. Moreover, these candidates are associated with false positives. For these, we suggest they could correspond to non-specific binders, which leaves room for further optimization of the PEPscan protocol. Another unexpected advance comes from the observation of the applicability of PEPscan for polysaccharides and labeled peptides, suggesting that PEPscan could become a large spectrum approach to investigate protein-binder interactions, the binder not necessarily being a protein.

Selective Disruption of Survivin's Protein‐Protein Interactions: A Supramolecular Approach Based on Guanidiniocarbonylpyrrole

Targeting specific protein binding sites to interfere with protein‐protein interactions (PPIs) is crucial for the rational modulation of biologically relevant processes. Survivin, which is highly overexpressed in most cancer cells and considered to be a key player of carcinogenesis, features two functionally relevant binding sites. Here, we demonstrate selective disruption of the Survivin/Histone H3 or the Survivin/Crm1 interaction using a supramolecular approach. By rational design we identified two structurally related ligands (L_NES and L_HIS), capable of selectively inhibiting these PPIs, leading to a reduction in cancer cell proliferation.

A random forest classifier for protein-protein docking models

Herein, we present the results of a machine learning approach we developed to single out correct 3D docking models of protein-protein complexes obtained by popular docking software. To this aim, we generated 3 × 10 4 docking models for each of the 230 complexes in the protein-protein benchmark, version 5, using three different docking programs (HADDOCK, FTDock and ZDOCK), for a cumulative set of ≈ 7 × 10 6 docking models. Three different machine learning approaches (Random Forest, Supported Vector Machine and Perceptron) were used to train classifiers with 158 different scoring functions (features). The Random Forest algorithm outperformed the other two algorithms and was selected for further optimization. Using a features selection algorithm, and optimizing the random forest hyperparameters, allowed us to train and validate a random forest classifier, named COnservation Driven Expert System (CoDES). Testing of CoDES on independent datasets, as well as results of its comparative performance with machine learning methods recently developed in the field for the scoring of docking decoys, confirm its state-of-the-art ability to discriminate correct from incorrect decoys both in terms of global parameters and in terms of decoys ranked at the top positions.

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Software and data availability statement

The docking models are available at https://doi.org/10.5281/zenodo.4012018. The programs underlying this article will be shared on request to the corresponding authors.

Protein-Protein Docking: Past, Present, and Future

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

Integrative Modeling of PROTAC-Mediated Ternary Complexes

Proteolysis-targeting chimeras (PROTACs), which selectively induce targeted protein degradation, represent an emerging drug discovery technology. Although numerous PROTACs have been reported, designing potent PROTACs still remains a great challenge, to some extent, due to insufficient structural data of Target-PROTAC-E3 ternary complexes. In this work, PROTAC-Model, an integrative computational method by combining the FRODOCK-based protocol and RosettaDock-based refinement, was developed to predict PROTAC-mediated ternary complex structures and tested on 14 cases. The quality of the models was evaluated using the criteria of the critical assessment of predicted interactions (CAPRI). Using the unbound structures, the FRODOCK-based protocol can generate the ternary complex structures with medium or high quality for 8 cases out of 14. With the refinement by RosettaDock, the cases with medium or high quality increase to 12. Compared with PRosettaC and the method developed by Drummond et al., PROTAC-Model shows better performance. In summary, PROTAC-Model should be useful for the rational design of PROTACs.

Distance-based reconstruction of protein quaternary structures from inter-chain contacts

Predicting the quaternary structure of protein complex is an important problem. Inter‐chain residue‐residue contact prediction can provide useful information to guide the ab initio reconstruction of quaternary structures. However, few methods have been developed to build quaternary structures from predicted inter‐chain contacts. Here, we develop the first method based on gradient descent optimization (GD) to build quaternary structures of protein dimers utilizing inter‐chain contacts as distance restraints. We evaluate GD on several datasets of homodimers and heterodimers using true/predicted contacts and monomer structures as input. GD consistently performs better than both simulated annealing and Markov Chain Monte Carlo simulation. Starting from an arbitrarily quaternary structure randomly initialized from the tertiary structures of protein chains and using true inter‐chain contacts as input, GD can reconstruct high‐quality structural models for homodimers and heterodimers with average TM‐score ranging from 0.92 to 0.99 and average interface root mean square distance from 0.72 Å to 1.64 Å. On a dataset of 115 homodimers, using predicted inter‐chain contacts as restraints, the average TM‐score of the structural models built by GD is 0.76. For 46% of the homodimers, high‐quality structural models with TM‐score ≥ 0.9 are reconstructed from predicted contacts. There is a strong correlation between the quality of the reconstructed models and the precision and recall of predicted contacts. Only a moderate precision or recall of inter‐chain contact prediction is needed to build good structural models for most homodimers. Moreover, GD improves the quality of quaternary structures predicted by AlphaFold2 on a Critical Assessment of Techniques for Protein Structure Prediction–Critical Assessments of Predictions of Interactions dataset.

A Novel In Silico Benchmarked Pipeline Capable of Complete Protein Analysis: A Possible Tool for Potential Drug Discovery

Simple Summary

Protein interactions govern the majority of an organism’s biological processes. Therefore, to fully understand the functionality of an organism, we must know how proteins work at a molecular level. This study assembled a protocol that enables scientists to construct a protein’s tertiary structure easily and subsequently to investigate its mechanism and function. Each step involved in prediction, validation, and functional analysis of a protein is crucial to obtain an accurate result. We have dubbed this the trifecta analysis. It was clear early in our research that no single study in the literature had previously encompassed the complete trifecta analysis. In particular, studies that recommend free, open-source tools that have been benchmarked for each step are lacking. The present study ensures that predictions are accurate and validated and will greatly benefit new and experienced scientists alike in obtaining a strong understanding of the trifecta analysis, resulting in a domino effect that could lead to drug development.

Abstract

Current in silico proteomics require the trifecta analysis, namely, prediction, validation, and functional assessment of a modeled protein. The main drawback of this endeavor is the lack of a single protocol that utilizes a proper set of benchmarked open-source tools to predict a protein’s structure and function accurately. The present study rectifies this drawback through the design and development of such a protocol. The protocol begins with the characterization of a novel coding sequence to identify the expressed protein. It then recognizes and isolates evolutionarily conserved sequence motifs through phylogenetics. The next step is to predict the protein’s secondary structure, followed by the prediction, refinement, and validation of its three-dimensional tertiary structure. These steps enable the functional analysis of the macromolecule through protein docking, which facilitates the identification of the protein’s active site. Each of these steps is crucial for the complete characterization of the protein under study. We have dubbed this process the trifecta analysis. In this study, we have proven the effectiveness of our protocol using the cystatin C and AChE proteins. Beginning with just their sequences, we have characterized both proteins’ structures and functions, including identifying the cystatin C protein’s seven-residue active site and the AChE protein’s active-site gorge via protein–protein and protein–ligand docking, respectively. This process will greatly benefit new and experienced scientists alike in obtaining a strong understanding of the trifecta analysis, resulting in a domino effect that could expand drug development.

D-SCRIPT translates genome to phenome with sequence-based, structure-aware, genome-scale predictions of protein-protein interactions

We combine advances in neural language modeling and structurally motivated design to develop D-SCRIPT, an interpretable and generalizable deep-learning model, which predicts interaction between two proteins using only their sequence and maintains high accuracy with limited training data and across species. We show that a D-SCRIPT model trained on 38,345 human PPIs enables significantly improved functional characterization of fly proteins compared with the state-of-the-art approach. Evaluating the same D-SCRIPT model on protein complexes with known 3D structure, we find that the inter-protein contact map output by D-SCRIPT has significant overlap with the ground truth. We apply D-SCRIPT to screen for PPIs in cow (Bos taurus) at a genome-wide scale and focusing on rumen physiology, identify functional gene modules related to metabolism and immune response. The predicted interactions can then be leveraged for function prediction at scale, addressing the genome-to-phenome challenge, especially in species where little data are available.

Conformational variability in proteins bound to single-stranded DNA: A new benchmark for new docking perspectives

We explored the Protein Data Bank (PDB) to collect protein-ssDNA structures and create a multi-conformational docking benchmark including both bound and unbound protein structures. Due to ssDNA high flexibility when not bound, no ssDNA unbound structure is included in the benchmark. For the 91 sequence-identity groups identified as bound-unbound structures of the same protein, we studied the conformational changes in the protein induced by the ssDNA binding. Moreover, based on several bound or unbound protein structures in some groups, we also assessed the intrinsic conformational variability in either bound or unbound conditions and compared it to the supposedly binding-induced modifications. To illustrate a use case of this benchmark, we performed docking experiments using ATTRACT docking software. This benchmark is, to our knowledge, the first one made to peruse available structures of ssDNA-protein interactions to such an extent, aiming to improve computational docking tools dedicated to this kind of molecular interactions.

DeepComplex: A Web Server of Predicting Protein Complex Structures by Deep Learning Inter-chain Contact Prediction and Distance-Based Modelling

Proteins interact to form complexes. Predicting the quaternary structure of protein complexes is useful for protein function analysis, protein engineering, and drug design. However, few user-friendly tools leveraging the latest deep learning technology for inter-chain contact prediction and the distance-based modelling to predict protein quaternary structures are available. To address this gap, we develop DeepComplex, a web server for predicting structures of dimeric protein complexes. It uses deep learning to predict inter-chain contacts in a homodimer or heterodimer. The predicted contacts are then used to construct a quaternary structure of the dimer by the distance-based modelling, which can be interactively viewed and analysed. The web server is freely accessible and requires no registration. It can be easily used by providing a job name and an email address along with the tertiary structure for one chain of a homodimer or two chains of a heterodimer. The output webpage provides the multiple sequence alignment, predicted inter-chain residue-residue contact map, and predicted quaternary structure of the dimer. DeepComplex web server is freely available at http://tulip.rnet.missouri.edu/deepcomplex/web_index.html.

Quality Assessment of Protein Docking Models Based on Graph Neural Network

Protein docking provides a structural basis for the design of drugs and vaccines. Among the processes of protein docking, quality assessment (QA) is utilized to pick near-native models from numerous protein docking candidate conformations, and it directly determines the final docking results. Although extensive efforts have been made to improve QA accuracy, it is still the bottleneck of current protein docking systems. In this paper, we presented a Deep Graph Attention Neural Network (DGANN) to evaluate and rank protein docking candidate models. DGANN learns inter-residue physio-chemical properties and structural fitness across the two protein monomers in a docking model and generates their probabilities of near-native models. On the ZDOCK decoy benchmark, our DGANN outperformed the ranking provided by ZDOCK in terms of ranking good models into the top selections. Furthermore, we conducted comparative experiments on an independent testing dataset, and the results also demonstrated the superiority and generalization of our proposed method.

Peptide-Based Inhibitors for SARS-CoV-2 and SARS-CoV

The COVID‐19 (coronavirus disease) global pandemic, caused by the spread of the SARS‐CoV‐2 (severe acute respiratory syndrome coronavirus 2) virus, currently has limited treatment options which include vaccines, anti‐virals, and repurposed therapeutics. With their high specificity, tunability, and biocompatibility, small molecules like peptides are positioned to act as key players in combating SARS‐CoV‐2, and can be readily modified to match viral mutation rate. A recent expansion of the understanding of the viral structure and entry mechanisms has led to the proliferation of therapeutic viral entry inhibitors. In this comprehensive review, inhibitors of SARS and SARS‐CoV‐2 are investigated and discussed based on therapeutic design, inhibitory mechanistic approaches, and common targets. Peptide therapeutics are highlighted, which have demonstrated in vitro or in vivo efficacy, discuss advantages of peptide therapeutics, and common strategies in identifying targets for viral inhibition.

Information-driven modeling of biomolecular complexes

Proteins play crucial roles in every cellular process by interacting with each other, nucleic acids, metabolites, and other molecules. The resulting assemblies can be very large and intricate and pose challenges to experimental methods. In the current era of integrative modeling, it is often only by a combination of various experimental techniques and computations that three-dimensional models of those molecular machines can be obtained. Among the various computational approaches available, molecular docking is often the method of choice when it comes to predicting three-dimensional structures of complexes. Docking can generate particularly accurate models when taking into account the available information on the complex of interest. We review here the use of experimental and bioinformatics data in protein-protein docking, describing recent software developments and highlighting applications for the modeling of antibody-antigen complexes and membrane protein complexes, and the use of evolutionary and shape information.

Information-Driven Docking for TCR-pMHC Complex Prediction

T cell receptor (TCR) recognition of peptides presented by major histocompatibility complex (MHC) molecules is a fundamental process in the adaptive immune system. An understanding of this recognition process at the molecular level is crucial for TCR based therapeutics and vaccine design. The broad nature of TCR diversity and cross-reactivity presents a challenge for traditional structural resolution. Computational modelling of TCR-pMHC complexes offers an efficient alternative. This study compares the ability of four general-purpose docking platforms (ClusPro, LightDock, ZDOCK and HADDOCK) to make use of varying levels of binding interface information for accurate TCR-pMHC modelling. Each platform was tested on an expanded benchmark set of 44 TCR-pMHC docking cases. In general, HADDOCK is shown to be the best performer. Docking strategy guidance is provided to obtain the best models for each platform for future research. The TCR-pMHC docking cases used in this study can be downloaded from https://github.com/innate2adaptive/ExpandedBenchmark.

Protein Docking Model Evaluation by Graph Neural Networks

Physical interactions of proteins play key functional roles in many important cellular processes. To understand molecular mechanisms of such functions, it is crucial to determine the structure of protein complexes. To complement experimental approaches, which usually take a considerable amount of time and resources, various computational methods have been developed for predicting the structures of protein complexes. In computational modeling, one of the challenges is to identify near-native structures from a large pool of generated models. Here, we developed a deep learning-based approach named Graph Neural Network-based DOcking decoy eValuation scorE (GNN-DOVE). To evaluate a protein docking model, GNN-DOVE extracts the interface area and represents it as a graph. The chemical properties of atoms and the inter-atom distances are used as features of nodes and edges in the graph, respectively. GNN-DOVE was trained, validated, and tested on docking models in the Dockground database and further tested on a combined dataset of Dockground and ZDOCK benchmark as well as a CAPRI scoring dataset. GNN-DOVE performed better than existing methods, including DOVE, which is our previous development that uses a convolutional neural network on voxelized structure models.

IsAb: a computational protocol for antibody design

The design of therapeutic antibodies has attracted a large amount of attention over the years. Antibodies are widely used to treat many diseases due to their high efficiency and low risk of adverse events. However, the experimental methods of antibody design are time-consuming and expensive. Although computational antibody design techniques have had significant advances in the past years, there are still some challenges that need to be solved, such as the flexibility of antigen structure, the lack of antibody structural data and the absence of standard antibody design protocol. In the present work, we elaborated on an in silico antibody design protocol for users to easily perform computer-aided antibody design. First, the Rosetta web server will be applied to generate the 3D structure of query antibodies if there is no structural information available. Then, two-step docking will be used to identify the binding pose of an antibody-antigen complex when the binding information is unknown. ClusPro is the first method to be used to conduct the global docking, and SnugDock is applied for the local docking. Sequentially, based on the predicted binding poses, in silico alanine scanning will be used to predict the potential hotspots (or key residues). Finally, computational affinity maturation protocol will be used to modify the structure of antibodies to theoretically increase their affinity and stability, which will be further validated by the bioassays in the future. As a proof of concept, we redesigned antibody D44.1 and compared it with previously reported data in order to validate IsAb protocol. To further illustrate our proposed protocol, we used cemiplimab antibody, a PD-1 checkpoint inhibitor, as an example to showcase a step-by-step tutorial.

Improved cluster ranking in protein-protein docking using a regression approach

We develop a Regression-based Ranking by Pairwise Cluster Comparisons (RRPCC) method to rank clusters of similar protein complex conformations generated by an underlying docking program. The method leverages robust regression to predict the relative quality difference between any pair or clusters and combines these pairwise assessments to form a ranked list of clusters, from higher to lower quality. We apply RRPCC to clusters produced by the automated docking server ClusPro and, depending on the training/validation strategy, we show improvement by 24-100% in ranking acceptable or better quality clusters first, and by 15-100% in ranking medium or better quality clusters first. We compare the RRPCC-ClusPro combination to a number of alternatives, and show that very different machine learning approaches to scoring docked structures yield similar success rates. Finally, we discuss the current limitations on sampling and scoring, looking ahead to further improvements. Interestingly, some features important for improved scoring are internal energy terms that occur only due to the local energy minimization applied in the refinement stage following rigid body docking.

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.

Decoding Protein-protein Interactions: An Overview

Drug discovery has focused on the paradigm "one drug, one target" for a long time. However, small molecules can act at multiple macromolecular targets, which serves as the basis for drug repurposing. In an effort to expand the target space, and given advances in X-ray crystallography, protein-protein interactions have become an emerging focus area of drug discovery enterprises. Proteins interact with other biomolecules and it is this intricate network of interactions that determines the behavior of the system and its biological processes. In this review, we briefly discuss networks in disease, followed by computational methods for protein-protein complex prediction. Computational methodologies and techniques employed towards objectives such as protein-protein docking, protein-protein interactions, and interface predictions are described extensively. Docking aims at producing a complex between proteins, while interface predictions identify a subset of residues on one protein that could interact with a partner, and protein-protein interaction sites address whether two proteins interact. In addition, approaches to predict hot spots and binding sites are presented along with a representative example of our internal project on the chemokine CXC receptor 3 B-isoform and predictive modeling with IP10 and PF4.

Hierarchical, rotation-equivariant neural networks to select structural models of protein complexes

Predicting the structure of multi-protein complexes is a grand challenge in biochemistry, with major implications for basic science and drug discovery. Computational structure prediction methods generally leverage predefined structural features to distinguish accurate structural models from less accurate ones. This raises the question of whether it is possible to learn characteristics of accurate models directly from atomic coordinates of protein complexes, with no prior assumptions. Here we introduce a machine learning method that learns directly from the 3D positions of all atoms to identify accurate models of protein complexes, without using any precomputed physics-inspired or statistical terms. Our neural network architecture combines multiple ingredients that together enable end-to-end learning from molecular structures containing tens of thousands of atoms: a point-based representation of atoms, equivariance with respect to rotation and translation, local convolutions, and hierarchical subsampling operations. When used in combination with previously developed scoring functions, our network substantially improves the identification of accurate structural models among a large set of possible models. Our network can also be used to predict the accuracy of a given structural model in absolute terms. The architecture we present is readily applicable to other tasks involving learning on 3D structures of large atomic systems.

A knowledge-based scoring function to assess quaternary associations of proteins

MOTIVATION

The elucidation of all inter-protein interactions would significantly enhance our knowledge of cellular processes at a molecular level. Given the enormity of the problem, the expenses and limitations of experimental methods, it is imperative that this problem is tackled computationally. In silico predictions of protein interactions entail sampling different conformations of the purported complex and then scoring these to assess for interaction viability. In this study, we have devised a new scheme for scoring protein-protein interactions.

RESULTS

Our method, PIZSA (Protein Interaction Z-Score Assessment), is a binary classification scheme for identification of native protein quaternary assemblies (binders/nonbinders) based on statistical potentials. The scoring scheme incorporates residue-residue contact preference on the interface with per residue-pair atomic contributions and accounts for clashes. PIZSA can accurately discriminate between native and non-native structural conformations from protein docking experiments and outperform other contact-based potential scoring functions. The method has been extensively benchmarked and is among the top 6 methods, outperforming 31 other statistical, physics based and machine learning scoring schemes. The PIZSA potentials can also distinguish crystallization artifacts from biological interactions.

AVAILABILITY AND IMPLEMENTATION

PIZSA is implemented as a web server at http://cospi.iiserpune.ac.in/pizsa and can be downloaded as a standalone package from http://cospi.iiserpune.ac.in/pizsa/Download/Download.html.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Evolution in non-peptide α-helix mimetics on the road to effective protein-protein interaction modulators

Modulation of interactome networks, essentially protein-protein interactions (PPIs), might represent valuable therapeutic approaches to different pathological conditions. Since a high percentage of PPIs are mediated by α-helical structures at the interacting surface, the development of compounds able to reproduce the amino acid side-chain organization of α-helices (e.g. stabilized α-helix peptides and β-derivatives, proteomimetics, and α-helix small-molecule mimetics) focuses the attention of different research groups. This appraisal describes the recent progress in the non-peptide α-helix mimetics field, which has evolved from single-face to multi-face reproducing compounds and from oligomeric to monomeric scaffolds able to bear different substituents in similar spatial dispositions as the side-chains in canonical helices. Grouped by chemical structures, the review contemplates terphenyl-like molecules, oligobenzamides and heterocyclic analogues, benzamide-amino acid conjugates and non-oligomeric small-molecules mimetics, among others, and their effectiveness to stabilize/disrupt therapeutically relevant PPIs. The X-ray structures of a couple of oligomeric peptidomimetics and of some small-molecules complexed with the MDM2 protein, as well as the state of the art on their development in clinical trials, are also remarked. The discovery of a continuously increasing number of new disease-relevant PPIs could offer future opportunities for these and other forthcoming α-helix mimetics.

Evaluation of CONSRANK-Like Scoring Functions for Rescoring Ensembles of Protein–Protein Docking Poses

Scoring is a challenging step in protein–protein docking, where typically thousands of solutions are generated. In this study, we ought to investigate the contribution of consensus-rescoring, as introduced by Oliva et al. (2013) with the CONSRANK method, where the set of solutions is used to build statistics in order to identify recurrent solutions. We explore several ways to perform consensus-based rescoring on the ZDOCK decoy set for Benchmark 4. We show that the information of the interface size is critical for successful rescoring in this context, but that consensus rescoring in itself performs less well than traditional physics-based evaluation. The results of physics-based and consensus-based rescoring are partially overlapping, supporting the use of a combination of these approaches.