Selecting machine-learning scoring functions for structure-based virtual screening

Inactive-enriched machine-learning models exploiting patent data improve structure-based virtual screening for PDL1 dimerizers

INTRODUCTION

Small-molecule Programmable Cell Death Protein 1/Programmable Death-Ligand 1 (PD1/PDL1) inhibition via PDL1 dimerization has the potential to lead to inexpensive drugs with better cancer patient outcomes and milder side effects. However, this therapeutic approach has proven challenging, with only one PDL1 dimerizer reaching early clinical trials so far. There is hence a need for fast and accurate methods to develop alternative PDL1 dimerizers.

OBJECTIVES

We aim to show that structure-based virtual screening (SBVS) based on PDL1-specific machine-learning (ML) scoring functions (SFs) is a powerful drug design tool for detecting PD1/PDL1 inhibitors via PDL1 dimerization.

METHODS

By incorporating the latest MLSF advances, we generated and evaluated PDL1-specific MLSFs (classifiers and inactive-enriched regressors) on two demanding test sets.

RESULTS

60 PDL1-specific MLSFs (30 classifiers and 30 regressors) were generated. Our large-scale analysis provides highly predictive PDL1-specific MLSFs that benefitted from training with large volumes of docked inactives and enabling inactive-enriched regression.

CONCLUSION

PDL1-specific MLSFs strongly outperformed generic SFs of various types on this target and are released here without restrictions.

Interaction-Based Inductive Bias in Graph Neural Networks: Enhancing Protein-Ligand Binding Affinity Predictions From 3D Structures

Inductive bias in machine learning (ML) is the set of assumptions describing how a model makes predictions. Different ML-based methods for protein-ligand binding affinity (PLA) prediction have different inductive biases, leading to different levels of generalization capability and interpretability. Intuitively, the inductive bias of an ML-based model for PLA prediction should fit in with biological mechanisms relevant for binding to achieve good predictions with meaningful reasons. To this end, we propose an interaction-based inductive bias to restrict neural networks to functions relevant for binding with two assumptions: 1) A protein-ligand complex can be naturally expressed as a heterogeneous graph with covalent and non-covalent interactions; 2) The predicted PLA is the sum of pairwise atom-atom affinities determined by non-covalent interactions. The interaction-based inductive bias is embodied by an explainable heterogeneous interaction graph neural network (EHIGN) for explicitly modeling pairwise atom-atom interactions to predict PLA from 3D structures. Extensive experiments demonstrate that EHIGN achieves better generalization capability than other state-of-the-art ML-based baselines in PLA prediction and structure-based virtual screening. More importantly, comprehensive analyses of distance-affinity, pose-affinity, and substructure-affinity relations suggest that the interaction-based inductive bias can guide the model to learn atomic interactions that are consistent with physical reality. As a case study to demonstrate practical usefulness, our method is tested for predicting the efficacy of Nirmatrelvir against SARS-CoV-2 variants. EHIGN successfully recognizes the changes in the efficacy of Nirmatrelvir for different SARS-CoV-2 variants with meaningful reasons.

Advances in molecular glues: exploring chemical space and design principles for targeted protein degradation

The discovery of the E3 ligase cereblon (CRBN) as the target of thalidomide and its analogs revolutionized the field of targeted protein degradation (TPD). This ubiquitin-mediated degradation pathway was first harnessed by bivalent degraders. Recently, the emergence of low-molecular-weight molecular glue degraders (MGDs) has expanded the TPD landscape, because MGDs operate via the same mechanism while offering attractive physicochemical properties that are consistent with small-molecule therapeutics. This review delves into the discovery and advancement of MGDs, with case studies on cyclin K and the zinc finger protein IKZF2, highlighting the design principles, biological assays and therapeutic applications. Additionally, it examines the chemical space of molecular glues and outlines the collaborative efforts that are fueling innovation in this field.

Data-centric challenges with the application and adoption of artificial intelligence for drug discovery

INTRODUCTION

Artificial intelligence (AI) is exhibiting tremendous potential to reduce the massive costs and long timescales of drug discovery. There are however important challenges currently limiting the impact and scope of AI models.

AREAS COVERED

In this perspective, the authors discuss a range of data issues (bias, inconsistency, skewness, irrelevance, small size, high dimensionality), how they challenge AI models, and which issue-specific mitigations have been effective. Next, they point out the challenges faced by uncertainty quantification techniques aimed at enhancing and trusting the predictions from these AI models. They also discuss how conceptual errors, unrealistic benchmarks and performance misestimation can confound the evaluation of models and thus their development. Lastly, the authors explain how human bias, whether from AI experts or drug discovery experts, constitutes another challenge that can be alleviated by gaining more prospective experience.

EXPERT OPINION

AI models are often developed to excel on retrospective benchmarks unlikely to anticipate their prospective performance. As a result, only a few of these models are ever reported to have prospective value (e.g. by discovering potent and innovative drug leads for a therapeutic target). The authors have discussed what can go wrong in practice with AI for drug discovery. The authors hope that this will help inform the decisions of editors, funders investors, and researchers working in this area.

Target Fisher: A Consensus Structure-Based Target Prediction Tool, and its Application in the Discovery of Selective MAO-B Inhibitors

In this work we introduce Target Fisher, a consensus structure-based target prediction tool that integrates molecular docking and machine learning with the aim to aid in the identification of potential biological targets and the optimization of the use of bioassays. Target Fisher uses per-residue energy decomposition profiles extracted from docking poses as fingerprints to train target-specific machine learning models. It provides predictions for a curated set of 37 protein targets, covering a diverse range of biological entities, and offers a user-friendly interface accessible via a web server (https://gqc.quimica.unlp.edu.ar/targetfisher/). In this sense, Target Fisher is a valuable tool to aid organic and medicinal chemistry groups in target identification, drug discovery and drug repurposing. As a case study, we demonstrate the efficacy of Target Fisher by screening a small library of assorted natural products for targets relevant to neurodegenerative diseases, which resulted in the identification and experimental validation of selective inhibitors of monoamine oxidase B (MAO-B).

Comprehensive machine learning boosts structure-based virtual screening for PARP1 inhibitors

Poly ADP-ribose polymerase 1 (PARP1) is an attractive therapeutic target for cancer treatment. Machine-learning scoring functions constitute a promising approach to discovering novel PARP1 inhibitors. Cutting-edge PARP1-specific machine-learning scoring functions were investigated using semi-synthetic training data from docking activity-labelled molecules: known PARP1 inhibitors, hard-to-discriminate decoys property-matched to them with generative graph neural networks and confirmed inactives. We further made test sets harder by including only molecules dissimilar to those in the training set. Comprehensive analysis of these datasets using five supervised learning algorithms, and protein-ligand fingerprints extracted from docking poses and ligand only features revealed one highly predictive scoring function. This is the PARP1-specific support vector machine-based regressor, when employing PLEC fingerprints, which achieved a high Normalized Enrichment Factor at the top 1% on the hardest test set (NEF1% = 0.588, median of 10 repetitions), and was more predictive than any other investigated scoring function, especially the classical scoring function employed as baseline.

The Impact of Data on Structure-Based Binding Affinity Predictions Using Deep Neural Networks

Artificial intelligence (AI) has gained significant traction in the field of drug discovery, with deep learning (DL) algorithms playing a crucial role in predicting protein-ligand binding affinities. Despite advancements in neural network architectures, system representation, and training techniques, the performance of DL affinity prediction has reached a plateau, prompting the question of whether it is truly solved or if the current performance is overly optimistic and reliant on biased, easily predictable data. Like other DL-related problems, this issue seems to stem from the training and test sets used when building the models. In this work, we investigate the impact of several parameters related to the input data on the performance of neural network affinity prediction models. Notably, we identify the size of the binding pocket as a critical factor influencing the performance of our statistical models; furthermore, it is more important to train a model with as much data as possible than to restrict the training to only high-quality datasets. Finally, we also confirm the bias in the typically used current test sets. Therefore, several types of evaluation and benchmarking are required to understand models' decision-making processes and accurately compare the performance of models.

"You've got the Body I've got the Brains" - Could the current AI-based tools replace the human ingenuity for designing new drug candidates?

The emergence of artificial intelligence (AI) tools has transformed the landscape of drug discovery, providing unprecedented speed, efficiency, and cost-effectiveness in the search for new therapeutics. From target identification to drug formulation and delivery, AI-driven algorithms have revolutionized various aspects of medicinal chemistry, significantly accelerating the drug design process. Despite the transformative power of AI, this perspective article emphasizes the limitations of AI tools in drug discovery, requiring inventive skills of medicinal chemists. However, the article highlighted that there is a need for a harmonious integration of AI-based tools and human expertise in drug discovery. Such a synergistic approach promises to lead to groundbreaking therapies that address unmet medical needs and benefit humankind. As the world evolves technologically, the question remains: When will AI tools effectively design and develop drugs? The answer may lie in the seamless collaboration between AI and human researchers, unlocking transformative therapies that combat diseases effectively.

The Impact of Supervised Learning Methods in Ultralarge High-Throughput Docking

Structure-based virtual screening methods are, nowadays, one of the key pillars of computational drug discovery. In recent years, a series of studies have reported docking-based virtual screening campaigns of large databases ranging from hundreds to thousands of millions compounds, further identifying novel hits after experimental validation. As these larg-scale efforts are not generally accessible, machine learning-based protocols have emerged to accelerate the identification of virtual hits within an ultralarge chemical space, reaching impressive reductions in computational time. Herein, we illustrate the motivation and the problem behind the screening of large databases, providing an overview of key concepts and essential applications of machine learning-accelerated protocols, specifically concerning supervised learning methods. We also discuss where the field stands with these novel developments, highlighting possible insights for future studies.

Beware of Simple Methods for Structure-Based Virtual Screening: The Critical Importance of Broader Comparisons

We discuss how data unbiasing and simple methods such as protein-ligand Interaction FingerPrint (IFP) can overestimate virtual screening performance. We also show that IFP is strongly outperformed by target-specific machine-learning scoring functions, which were not considered in a recent report concluding that simple methods were better than machine-learning scoring functions at virtual screening.

On the Best Way to Cluster NCI-60 Molecules

Machine learning-based models have been widely used in the early drug-design pipeline. To validate these models, cross-validation strategies have been employed, including those using clustering of molecules in terms of their chemical structures. However, the poor clustering of compounds will compromise such validation, especially on test molecules dissimilar to those in the training set. This study aims at finding the best way to cluster the molecules screened by the National Cancer Institute (NCI)-60 project by comparing hierarchical, Taylor-Butina, and uniform manifold approximation and projection (UMAP) clustering methods. The best-performing algorithm can then be used to generate clusters for model validation strategies. This study also aims at measuring the impact of removing outlier molecules prior to the clustering step. Clustering results are evaluated using three well-known clustering quality metrics. In addition, we compute an average similarity matrix to assess the quality of each cluster. The results show variation in clustering quality from method to method. The clusters obtained by the hierarchical and Taylor-Butina methods are more computationally expensive to use in cross-validation strategies, and both cluster the molecules poorly. In contrast, the UMAP method provides the best quality, and therefore we recommend it to analyze this highly valuable dataset.

Advances in the discovery of new chemotypes through ultra-large library docking

INTRODUCTION

The size and complexity of virtual screening libraries in drug discovery have skyrocketed in recent years, reaching up to multiple billions of accessible compounds. However, virtual screening of such ultra-large libraries poses several challenges associated with preparing the libraries, sampling, and pre-selection of suitable compounds. The utilization of artificial intelligence (AI)-assisted screening approaches, such as deep learning, poses a promising countermeasure to deal with this rapidly expanding chemical space. For example, various AI-driven methods were recently successfully used to identify novel small molecule inhibitors of the SARS-CoV-2 main protease (M^pro).

AREAS COVERED

This review focuses on presenting various kinds of virtual screening methods suitable for dealing with ultra-large libraries. Challenges associated with these computational methodologies are discussed, and recent advances are highlighted in the example of the discovery of novel M^pro inhibitors targeting the SARS-CoV-2 virus.

EXPERT OPINION

With the rapid expansion of the virtual chemical space, the methodologies for docking and screening such quantities of molecules need to keep pace. Employment of AI-driven screening compounds has already been shown to be effective in a range from a few thousand to multiple billion compounds, furthered by de novo generation of drug-like molecules without human interference.

A fully differentiable ligand pose optimization framework guided by deep learning and a traditional scoring function

The recently reported machine learning- or deep learning-based scoring functions (SFs) have shown exciting performance in predicting protein-ligand binding affinities with fruitful application prospects. However, the differentiation between highly similar ligand conformations, including the native binding pose (the global energy minimum state), remains challenging that could greatly enhance the docking. In this work, we propose a fully differentiable, end-to-end framework for ligand pose optimization based on a hybrid SF called DeepRMSD+Vina combined with a multi-layer perceptron (DeepRMSD) and the traditional AutoDock Vina SF. The DeepRMSD+Vina, which combines (1) the root mean square deviation (RMSD) of the docking pose with respect to the native pose and (2) the AutoDock Vina score, is fully differentiable; thus is capable of optimizing the ligand binding pose to the energy-lowest conformation. Evaluated by the CASF-2016 docking power dataset, the DeepRMSD+Vina reaches a success rate of 94.4%, which outperforms most reported SFs to date. We evaluated the ligand conformation optimization framework in practical molecular docking scenarios (redocking and cross-docking tasks), revealing the high potentialities of this framework in drug design and discovery. Structural analysis shows that this framework has the ability to identify key physical interactions in protein-ligand binding, such as hydrogen-bonding. Our work provides a paradigm for optimizing ligand conformations based on deep learning algorithms. The DeepRMSD+Vina model and the optimization framework are available at GitHub repository https://github.com/zchwang/DeepRMSD-Vina_Optimization.

Reducing false positive rate of docking-based virtual screening by active learning

Machine learning-based scoring functions (MLSFs) have become a very favorable alternative to classical scoring functions because of their potential superior screening performance. However, the information of negative data used to construct MLSFs was rarely reported in the literature, and meanwhile the putative inactive molecules recorded in existing databases usually have obvious bias from active molecules. Here we proposed an easy-to-use method named AMLSF that combines active learning using negative molecular selection strategies with MLSF, which can iteratively improve the quality of inactive sets and thus reduce the false positive rate of virtual screening. We chose energy auxiliary terms learning as the MLSF and validated our method on eight targets in the diverse subset of DUD-E. For each target, we screened the IterBioScreen database by AMLSF and compared the screening results with those of the four control models. The results illustrate that the number of active molecules in the top 1000 molecules identified by AMLSF was significantly higher than those identified by the control models. In addition, the free energy calculation results for the top 10 molecules screened out by the AMLSF, null model and control models based on DUD-E also proved that more active molecules can be identified, and the false positive rate can be reduced by AMLSF.

Designing drugs when there is low data availability: one-shot learning and other approaches to face the issues of a long-term concern

INTRODUCTION

Modern drug discovery generally is accessed by useful information from previous large databases or uncovering novel data. The lack of biological and/or chemical data tends to slow the development of scientific research and innovation. Here, approaches that may help provide solutions to generate or obtain enough relevant data or improve/accelerate existing methods within the last five years were reviewed.

AREAS COVERED

One-shot learning (OSL) approaches, structural modeling, molecular docking, scoring function space (SFS), molecular dynamics (MD), and quantum mechanics (QM) may be used to amplify the amount of available data to drug design and discovery campaigns, presenting methods, their perspectives, and discussions to be employed in the near future.

EXPERT OPINION

Recent works have successfully used these techniques to solve a range of issues in the face of data scarcity, including complex problems such as the challenging scenario of drug design aimed at intrinsically disordered proteins and the evaluation of potential adverse effects in a clinical scenario. These examples show that it is possible to improve and kickstart research from scarce available data to design and discover new potential drugs.

SCORCH: Improving structure-based virtual screening with machine learning classifiers, data augmentation, and uncertainty estimation

INTRODUCTION

The discovery of a new drug is a costly and lengthy endeavour. The computational prediction of which small molecules can bind to a protein target can accelerate this process if the predictions are fast and accurate enough. Recent machine-learning scoring functions re-evaluate the output of molecular docking to achieve more accurate predictions. However, previous scoring functions were trained on crystalised protein-ligand complexes and datasets of decoys. The limited availability of crystal structures and biases in the decoy datasets can lower the performance of scoring functions.

OBJECTIVES

To address key limitations of previous scoring functions and thus improve the predictive performance of structure-based virtual screening.

METHODS

A novel machine-learning scoring function was created, named SCORCH (Scoring COnsensus for RMSD-based Classification of Hits). To develop SCORCH, training data is augmented by considering multiple ligand poses and labelling poses based on their RMSD from the native pose. Decoy bias is addressed by generating property-matched decoys for each ligand and using the same methodology for preparing and docking decoys and ligands. A consensus of 3 different machine learning approaches is also used to improve performance.

RESULTS

We find that multi-pose augmentation in SCORCH improves its docking power and screening power on independent benchmark datasets. SCORCH outperforms an equivalent scoring function trained on single poses, with a 1% enrichment factor (EF) of 13.78 vs. 10.86 on 18 DEKOIS 2.0 targets and a mean native pose rank of 5.9 vs 30.4 on CSAR 2014. Additionally, SCORCH outperforms widely used scoring functions in virtual screening and pose prediction on independent benchmark datasets.

CONCLUSION

By rationally addressing key limitations of previous scoring functions, SCORCH improves the performance of virtual screening. SCORCH also provides an estimate of its uncertainty, which can help reduce the cost and time required for drug discovery.

Structure-based drug repurposing: traditional and advanced AI/ML-aided methods

The current global health emergency in the form of the Coronavirus 2019 (COVID-19) pandemic has highlighted the need for fast, accurate, and efficient drug discovery pipelines. Traditional drug discovery projects relying on in vitro high-throughput screening (HTS) involve large investments and sophisticated experimental set-ups, affordable only to big biopharmaceutical companies. In this scenario, application of efficient state-of-the-art computational methods and modern artificial intelligence (AI)-based algorithms for rapid screening of repurposable chemical space [approved drugs and natural products (NPs) with proven pharmacokinetic profiles] to identify the initial leads is a powerful option to save resources and time. Structure-based drug repurposing is a popular in silico repurposing approach. In this review, we discuss traditional and modern AI-based computational methods and tools applied at various stages for structure-based drug discovery (SBDD) pipelines. Additionally, we highlight the role of generative models in generating molecules with scaffolds from repurposable chemical space. Teaser: This review highlights the importance of repurposable chemical space, and the contributions of conventional in silico approaches and modern machine-learning algorithms for rapid structure-based drug repurposing.

Efficient Hit-to-Lead Searching of Kinase Inhibitor Chemical Space via Computational Fragment Merging

In early-stage drug discovery, the hit-to-lead optimization (or "hit expansion") stage entails starting from a newly identified active compound and improving its potency or other properties. Traditionally, this process relies on synthesizing and evaluating a series of analogues to build up structure-activity relationships. Here, we describe a computational strategy focused on kinase inhibitors, intended to expedite the process of identifying analogues with improved potency. Our protocol begins from an inhibitor of the target kinase and generalizes the synthetic route used to access it. By searching for commercially available replacements for the individual building blocks used to make the parent inhibitor, we compile an enumerated library of compounds that can be accessed using the same chemical transformations; these huge libraries can exceed many millions─or billions─of compounds. Because the resulting libraries are much too large for explicit virtual screening, we instead consider alternate approaches to identify the top-scoring compounds. We find that contributions from individual substituents are well described by a pairwise additivity approximation, provided that the corresponding fragments position their shared core in precisely the same way relative to the binding site. This key insight allows us to determine which fragments are suitable for merging into single new compounds and which are not. Further, the use of pairwise approximation allows interaction energies to be assigned to each compound in the library without the need for any further structure-based modeling: interaction energies instead can be reliably estimated from the energies of the component fragments, and the reduced computational requirements allow for flexible energy minimizations that allow the kinase to respond to each substitution. We demonstrate this protocol using libraries built from six representative kinase inhibitors drawn from the literature, which target five different kinases: CDK9, CHK1, CDK2, EGFR^T790M, and ACK1. In each example, the enumerated library includes additional analogues reported by the original study to have activity, and these analogues are successfully prioritized within the library. We envision that the insights from this work can facilitate the rapid assembly and screening of increasingly large libraries for focused hit-to-lead optimization. To enable adoption of these methods and to encourage further analyses, we disseminate the computational tools needed to deploy this protocol.

CSM-carbohydrate: protein-carbohydrate binding affinity prediction and docking scoring function

Protein-carbohydrate interactions are crucial for many cellular processes but can be challenging to biologically characterise. To improve our understanding and ability to model these molecular interactions, we used a carefully curated set of 370 protein-carbohydrate complexes with experimental structural and biophysical data in order to train and validate a new tool, cutoff scanning matrix (CSM)-carbohydrate, using machine learning algorithms to accurately predict their binding affinity and rank docking poses as a scoring function. Information on both protein and carbohydrate complementarity, in terms of shape and chemistry, was captured using graph-based structural signatures. Across both training and independent test sets, we achieved comparable Pearson's correlations of 0.72 under cross-validation [root mean square error (RMSE) of 1.58 Kcal/mol] and 0.67 on the independent test (RMSE of 1.72 Kcal/mol), providing confidence in the generalisability and robustness of the final model. Similar performance was obtained across mono-, di- and oligosaccharides, further highlighting the applicability of this approach to the study of larger complexes. We show CSM-carbohydrate significantly outperformed previous approaches and have implemented our method and make all data freely available through both a user-friendly web interface and application programming interface, to facilitate programmatic access at http://biosig.unimelb.edu.au/csm_carbohydrate/. We believe CSM-carbohydrate will be an invaluable tool for helping assess docking poses and the effects of mutations on protein-carbohydrate affinity, unravelling important aspects that drive binding recognition.

Improving Structure-Based Virtual Screening with Ensemble Docking and Machine Learning

One of the main challenges of structure-based virtual screening (SBVS) is the incorporation of the receptor's flexibility, as its explicit representation in every docking run implies a high computational cost. Therefore, a common alternative to include the receptor's flexibility is the approach known as ensemble docking. Ensemble docking consists of using a set of receptor conformations and performing the docking assays over each of them. However, there is still no agreement on how to combine the ensemble docking results to obtain the final ligand ranking. A common choice is to use consensus strategies to aggregate the ensemble docking scores, but these strategies exhibit slight improvement regarding the single-structure approach. Here, we claim that using machine learning (ML) methodologies over the ensemble docking results could improve the predictive power of SBVS. To test this hypothesis, four proteins were selected as study cases: CDK2, FXa, EGFR, and HSP90. Protein conformational ensembles were built from crystallographic structures, whereas the evaluated compound library comprised up to three benchmarking data sets (DUD, DEKOIS 2.0, and CSAR-2012) and cocrystallized molecules. Ensemble docking results were processed through 30 repetitions of 4-fold cross-validation to train and validate two ML classifiers: logistic regression and gradient boosting trees. Our results indicate that the ML classifiers significantly outperform traditional consensus strategies and even the best performance case achieved with single-structure docking. We provide statistical evidence that supports the effectiveness of ML to improve the ensemble docking performance.

Combination of pose and rank consensus in docking-based virtual screening: the best of both worlds

The use of high-throughput docking (HTD) in the drug discovery pipeline is today widely established. In spite of methodological improvements in docking accuracy (pose prediction), scoring power, ranking power, and screening power in HTD remain challenging. In fact, pose prediction is of critical importance in view of the pose-dependent scoring process, since incorrect poses will necessarily decrease the ranking power of scoring functions. The combination of results from different docking programs (consensus scoring) has been shown to improve the performance of HTD. Moreover, it has been also shown that a pose consensus approach might also result in database enrichment. We present a new methodology named Pose/Ranking Consensus (PRC) that combines both pose and ranking consensus approaches, to overcome the limitations of each stand-alone strategy. This approach has been developed using four docking programs (ICM, rDock, Auto Dock 4, and PLANTS; the first one is commercial, the other three are free). We undertook a thorough analysis for the best way of combining pose and rank strategies, and applied the PRC to a wide range of 34 targets sampling different protein families and binding site properties. Our approach exhibits an improved systematic performance in terms of enrichment factor and hit rate with respect to either pose consensus or consensus ranking alone strategies at a lower computational cost, while always ensuring the recovery of a suitable number of ligands. An analysis using four free docking programs (replacing ICM by Auto Dock Vina) displayed comparable results.

The new methodology named Pose/Ranking Consensus (PRC) combines both pose and ranking consensus strategies. It displays an enhanced performance in terms of enrichment factor and hit rate, ensuring the recovery of a suitable number of ligands.

The impact of cross-docked poses on performance of machine learning classifier for protein-ligand binding pose prediction

Structure-based drug design depends on the detailed knowledge of the three-dimensional (3D) structures of protein-ligand binding complexes, but accurate prediction of ligand-binding poses is still a major challenge for molecular docking due to deficiency of scoring functions (SFs) and ignorance of protein flexibility upon ligand binding. In this study, based on a cross-docking dataset dedicatedly constructed from the PDBbind database, we developed several XGBoost-trained classifiers to discriminate the near-native binding poses from decoys, and systematically assessed their performance with/without the involvement of the cross-docked poses in the training/test sets. The calculation results illustrate that using Extended Connectivity Interaction Features (ECIF), Vina energy terms and docking pose ranks as the features can achieve the best performance, according to the validation through the random splitting or refined-core splitting and the testing on the re-docked or cross-docked poses. Besides, it is found that, despite the significant decrease of the performance for the threefold clustered cross-validation, the inclusion of the Vina energy terms can effectively ensure the lower limit of the performance of the models and thus improve their generalization capability. Furthermore, our calculation results also highlight the importance of the incorporation of the cross-docked poses into the training of the SFs with wide application domain and high robustness for binding pose prediction. The source code and the newly-developed cross-docking datasets can be freely available at https://github.com/sc8668/ml_pose_prediction and https://zenodo.org/record/5525936 , respectively, under an open-source license. We believe that our study may provide valuable guidance for the development and assessment of new machine learning-based SFs (MLSFs) for the predictions of protein-ligand binding poses.

Advance in dietary polyphenols as dipeptidyl peptidase-IV inhibitors to alleviate type 2 diabetes mellitus: aspects from structure-activity relationship and characterization methods

Dietary polyphenols with great antidiabetic effects are the most abundant components in edible products. Dietary polyphenols have attracted attention as dipeptidyl peptidase-IV (DPP-IV) inhibitors and indirectly improve insulin secretion. The DPP-IV inhibitory activities of dietary polyphenols depend on their structural diversity. Screening methods that can be used to rapidly and accurately identify potential polyphenol DPP-IV inhibitors are urgently needed. This review focuses on the relationship between the structures of dietary polyphenols and their DPP-IV inhibitory effects. Different characterization methods used for polyphenols as DPP-IV inhibitors have been summarized and compared. We conclude that the position and number of hydroxyl groups, methoxy groups, glycosylated groups, and the extent of conjugation influence the efficiency of inhibition of DPP-IV. Various combinations of methods, such as in-vitro enzymatic inhibition, ex-vivo/in-vivo enzymatic inhibition, cell-based in situ, and in-silico virtual screening, are used to evaluate the DPP-IV inhibitory effects of dietary polyphenols. Further investigations of polyphenol DPP-IV inhibitors will improve the bioaccessibility and bioavailability of these bioactive compounds. Exploration of (i) dietary polyphenols derived from multiple targets, that can prevent diabetes, and (ii) actual binding interactions via multispectral analysis, to understand the binding interactions in the complexes, is required.

Success stories of AI in drug discovery - where do things stand?

INTRODUCTION

Artificial intelligence (AI) in drug discovery and development (DDD) has gained more traction in the past few years. Many scientific reviews have already been made available in this area. Thus, in this review, the authors have focused on the success stories of AI-driven drug candidates and the scientometric analysis of the literature in this field.

AREA COVERED

The authors explore the literature to compile the success stories of AI-driven drug candidates that are currently being assessed in clinical trials or have investigational new drug (IND) status. The authors also provide the reader with their expert perspectives for future developments and their opinions on the field.

EXPERT OPINION

Partnerships between AI companies and the pharma industry are booming. The early signs of the impact of AI on DDD are encouraging, and the pharma industry is hoping for breakthroughs. AI can be a promising technology to unveil the greatest successes, but it has yet to be proven as AI is still at the embryonic stage.

Structure-based in silico approaches for drug discovery against Mycobacterium tuberculosis

Mycobacterium tuberculosis is the causative agent of TB and was estimated to cause 1.4 million death in 2019, alongside 10 million new infections. Drug resistance is a growing issue, with multi-drug resistant infections representing 3.3% of all new infections, hence novel antimycobacterial drugs are urgently required to combat this growing health emergency. Alongside this, increased knowledge of gene essentiality in the pathogenic organism and larger compound databases can aid in the discovery of new drug compounds. The number of protein structures, X-ray based and modelled, is increasing and now accounts for greater than > 80% of all predicted M. tuberculosis proteins; allowing novel targets to be investigated. This review will focus on structure-based in silico approaches for drug discovery, covering a range of complexities and computational demands, with associated antimycobacterial examples. This includes molecular docking, molecular dynamic simulations, ensemble docking and free energy calculations. Applications of machine learning onto each of these approaches will be discussed. The need for experimental validation of computational hits is an essential component, which is unfortunately missing from many current studies. The future outlooks of these approaches will also be discussed.

Deep Scoring Neural Network Replacing the Scoring Function Components to Improve the Performance of Structure-Based Molecular Docking

Accurate prediction of protein-ligand interactions can greatly promote drug development. Recently, a number of deep-learning-based methods have been proposed to predict protein-ligand binding affinities. However, these methods independently extract the feature representations of proteins and ligands but ignore the relative spatial positions and interaction pairs between them. Here, we propose a virtual screening method based on deep learning, called Deep Scoring, which directly extracts the relative position information and atomic attribute information on proteins and ligands from the docking poses. Furthermore, we use two Resnets to extract the features of ligand atoms and protein residues, respectively, and generate an atom-residue interaction matrix to learn the underlying principles of the interactions between proteins and ligands. This is then followed by a dual attention network (DAN) to generate the attention for two related entities (i.e., proteins and ligands) and to weigh the contributions of each atom and residue to binding affinity prediction. As a result, Deep Scoring outperforms other structure-based deep learning methods in terms of screening performance (area under the receiver operating characteristic curve (AUC) of 0.901 for an unbiased DUD-E version), pose prediction (AUC of 0.935 for PDBbind test set), and generalization ability (AUC of 0.803 for the CHEMBL data set). Finally, Deep Scoring was used to select novel ERK2 inhibitor, and two compounds (D264-0698 and D483-1785) were obtained with potential inhibitory activity on ERK2 through the biological experiments.

Computational representations of protein-ligand interfaces for structure-based virtual screening

Introduction: Structure-based virtual screening (SBVS) is an essential strategy for hit identification. SBVS primarily uses molecular docking, which exploits the protein-ligand binding mode and associated affinity score for compound ranking. Previous studies have shown that computational representation of protein-ligand interfaces and the later establishment of machine learning models are efficacious in improving the accuracy of SBVS.Areas covered: The authors review the computational methods for representing protein-ligand interfaces, which include the traditional ones that use deliberately designed fingerprints and descriptors and the more recent methods that automatically extract features with deep learning. The effects of these methods on the performance of machine learning models are briefly discussed. Additionally, case studies that applied various computational representations to machine learning are cited with remarks.Expert opinion: It has become a trend to extract binding features automatically by deep learning, which uses a completely end-to-end representation. However, there is still plenty of scope for improvement . The interpretability of deep-learning models, the organization of data management, the quantity and quality of available data, and the optimization of hyperparameters could impact the accuracy of feature extraction. In addition, other important structural factors such as water molecules and protein flexibility should be considered.

Resources and computational strategies to advance small molecule SARS-CoV-2 discovery: Lessons from the pandemic and preparing for future health crises

There is an urgent need to identify new therapies that prevent SARS-CoV-2 infection and improve the outcome of COVID-19 patients. This pandemic has thus spurred intensive research in most scientific areas and in a short period of time, several vaccines have been developed. But, while the race to find vaccines for COVID-19 has dominated the headlines, other types of therapeutic agents are being developed. In this mini-review, we report several databases and online tools that could assist the discovery of anti-SARS-CoV-2 small chemical compounds and peptides. We then give examples of studies that combined in silico and in vitro screening, either for drug repositioning purposes or to search for novel bioactive compounds. Finally, we question the overall lack of discussion and plan observed in academic research in many countries during this crisis and suggest that there is room for improvement.

Artificial intelligence in drug discovery: recent advances and future perspectives

Introduction: Artificial intelligence (AI) has inspired computer-aided drug discovery. The widespread adoption of machine learning, in particular deep learning, in multiple scientific disciplines, and the advances in computing hardware and software, among other factors, continue to fuel this development. Much of the initial skepticism regarding applications of AI in pharmaceutical discovery has started to vanish, consequently benefitting medicinal chemistry.Areas covered: The current status of AI in chemoinformatics is reviewed. The topics discussed herein include quantitative structure-activity/property relationship and structure-based modeling, de novo molecular design, and chemical synthesis prediction. Advantages and limitations of current deep learning applications are highlighted, together with a perspective on next-generation AI for drug discovery.Expert opinion: Deep learning-based approaches have only begun to address some fundamental problems in drug discovery. Certain methodological advances, such as message-passing models, spatial-symmetry-preserving networks, hybrid de novo design, and other innovative machine learning paradigms, will likely become commonplace and help address some of the most challenging questions. Open data sharing and model development will play a central role in the advancement of drug discovery with AI.