Efficient Exploration of Chemical Space with Docking and Deep Learning

Thompson Sampling─An Efficient Method for Searching Ultralarge Synthesis on Demand Databases

Over the last five years, virtual screening of ultralarge synthesis on-demand libraries has emerged as a powerful tool for hit identification in drug discovery programs. As these libraries have grown to tens of billions of molecules, we have reached a point where it is no longer cost-effective to screen every molecule virtually. To address these challenges, several groups have developed heuristic search methods to rapidly identify the best molecules on a virtual screen. This article describes the application of Thompson sampling (TS), an active learning approach that streamlines the virtual screening of large combinatorial libraries by performing a probabilistic search in the reagent space, thereby never requiring the full enumeration of the library. TS is a general technique that can be applied to various virtual screening modalities, including 2D and 3D similarity search, docking, and application of machine-learning models. In an illustrative example, we show that TS can identify more than half of the top 100 molecules from a docking-based virtual screen of 335 million molecules by evaluating 1% of the data set.

Perspectives on Computational Enzyme Modeling: From Mechanisms to Design and Drug Development

Understanding enzyme mechanisms is essential for unraveling the complex molecular machinery of life. In this review, we survey the field of computational enzymology, highlighting key principles governing enzyme mechanisms and discussing ongoing challenges and promising advances. Over the years, computer simulations have become indispensable in the study of enzyme mechanisms, with the integration of experimental and computational exploration now established as a holistic approach to gain deep insights into enzymatic catalysis. Numerous studies have demonstrated the power of computer simulations in characterizing reaction pathways, transition states, substrate selectivity, product distribution, and dynamic conformational changes for various enzymes. Nevertheless, significant challenges remain in investigating the mechanisms of complex multistep reactions, large-scale conformational changes, and allosteric regulation. Beyond mechanistic studies, computational enzyme modeling has emerged as an essential tool for computer-aided enzyme design and the rational discovery of covalent drugs for targeted therapies. Overall, enzyme design/engineering and covalent drug development can greatly benefit from our understanding of the detailed mechanisms of enzymes, such as protein dynamics, entropy contributions, and allostery, as revealed by computational studies. Such a convergence of different research approaches is expected to continue, creating synergies in enzyme research. This review, by outlining the ever-expanding field of enzyme research, aims to provide guidance for future research directions and facilitate new developments in this important and evolving field.

12-(S)-Hydroxyeicosatetraenoic Acid and GPR31 Signaling in Spinal Cord in Neuropathic Pain

Neuropathic pain is a pressing unmet medical need requiring novel nonopioid-based therapeutic approaches. Using unbiased transcriptomic analysis, we found that the expression of Gpr31, a G protein-coupled receptor, increased in the dorsal horn of the spinal cord in rats with traumatic nerve injury-induced neuropathic pain. Daily intrathecal injections of siGpr31 reversed behavioral hypersensitivities in a time-dependent manner. GPR31, a Gα i protein-coupled receptor, has recently been cloned and is a receptor for 12-(S)-hydroxyeicosatetraenoic acid [12-(S)-HETE]. The lack of commercially available GPR31 antagonists has hampered the understanding of this receptor in pathophysiological states, including pain. To investigate this, our first approach was to identify novel GPR31 antagonists. Using a multidisciplinary approach, including in silico modeling, we identified the first highly potent and selective small-molecule GPR31 antagonist, SAH2. Here, we characterize the pharmacological activity in well-described models of neuropathic pain in rodents and provide evidence that 12-(S)-HETE/GPR31-dependent behavioral hypersensitivities are mediated through mitogen-activated protein kinase (MAPK) activation in the spinal cord. Our studies provide the pharmacological rationale for investigating contributions of GPR31 along the pain neuroaxis and the development of nonopioid GPR31-targeted strategies. SIGNIFICANCE STATEMENT: We have identified the first highly selective GPR31 antagonist. Using this antagonist, we have demonstrated that GPR31 signaling in the spinal cord is pronociceptive and MAPK pathways provided signaling mechanisms downstream of GPR31 activation in these processes.

Integrating QSAR modelling and deep learning in drug discovery: the emergence of deep QSAR

Quantitative structure-activity relationship (QSAR) modelling, an approach that was introduced 60 years ago, is widely used in computer-aided drug design. In recent years, progress in artificial intelligence techniques, such as deep learning, the rapid growth of databases of molecules for virtual screening and dramatic improvements in computational power have supported the emergence of a new field of QSAR applications that we term 'deep QSAR'. Marking a decade from the pioneering applications of deep QSAR to tasks involved in small-molecule drug discovery, we herein describe key advances in the field, including deep generative and reinforcement learning approaches in molecular design, deep learning models for synthetic planning and the application of deep QSAR models in structure-based virtual screening. We also reflect on the emergence of quantum computing, which promises to further accelerate deep QSAR applications and the need for open-source and democratized resources to support computer-aided drug design.

Active Learning Approach for Guiding Site-of-Metabolism Measurement and Annotation

The ability to determine and predict metabolically labile atom positions in a molecule (also called "sites of metabolism" or "SoMs") is of high interest to the design and optimization of bioactive compounds, such as drugs, agrochemicals, and cosmetics. In recent years, several in silico models for SoM prediction have become available, many of which include a machine-learning component. The bottleneck in advancing these approaches is the coverage of distinct atom environments and rare and complex biotransformation events with high-quality experimental data. Pharmaceutical companies typically have measured metabolism data available for several hundred to several thousand compounds. However, even for metabolism experts, interpreting these data and assigning SoMs are challenging and time-consuming. Therefore, a significant proportion of the potential of the existing metabolism data, particularly in machine learning, remains dormant. Here, we report on the development and validation of an active learning approach that identifies the most informative atoms across molecular data sets for SoM annotation. The active learning approach, built on a highly efficient reimplementation of SoM predictor FAME 3, enables experts to prioritize their SoM experimental measurements and annotation efforts on the most rewarding atom environments. We show that this active learning approach yields competitive SoM predictors while requiring the annotation of only 20% of the atom positions required by FAME 3. The source code of the approach presented in this work is publicly available.

Docking for Molecules That Bind in a Symmetric Stack with SymDOCK

Discovering ligands for amyloid fibrils, such as those formed by the tau protein, is an area of great current interest. In recent structures, ligands bind in stacks in the tau fibrils to reflect the rotational and translational symmetry of the fibril itself; in these structures, the ligands make few interactions with the protein but interact extensively with each other. To exploit this symmetry and stacking, we developed SymDOCK, a method to dock molecules that follow the protein's symmetry. For each prospective ligand pose, we apply the symmetry operation of the fibril to generate a self-interacting and fibril-interacting stack, checking that doing so will not cause a clash between the original molecule and its image. Absent a clash, we retain that pose and add the ligand-ligand van der Waals energy to the ligand's docking score (here using DOCK3.8). We can check these geometries and energies using an implementation of ANI, a neural-network-based quantum-mechanical evaluation of the ligand stacking energies. In retrospective calculations, symmetry docking can reproduce the poses of three tau PET tracers whose structures have been determined. More convincingly, in a prospective study, SymDOCK predicted the structure of the PET tracer MK-6240 bound in a symmetrical stack to AD PHF tau before that structure was determined; the docked pose was used to determine how MK-6240 fit the cryo-EM density. In proof-of-concept studies, SymDOCK enriched known ligands over property-matched decoys in retrospective screens without sacrificing docking speed and can address large library screens that seek new symmetrical stackers. Future applications of this approach will be considered.

Considerations Around Structure-Based Drug Discovery for KRAS Using DOCK

Molecular docking is a popular computational tool in drug discovery. Leveraging structural information, docking software predicts binding poses of small molecules to cavities on the surfaces of proteins. Virtual screening for ligand discovery is a useful application of docking software. In this chapter, using the enigmatic KRAS protein as an example system, we endeavor to teach the reader about best practices for performing molecular docking with UCSF DOCK. We discuss methods for virtual screening and docking molecules on KRAS. We present the following six points to optimize our docking setup for prosecuting a virtual screen: protein structure choice, pocket selection, optimization of the scoring function, modification of sampling spheres and sampling procedures, choosing an appropriate portion of chemical space to dock, and the choice of which top scoring molecules to pick for purchase.

Integration of Computational Docking into Anti-Cancer Drug Response Prediction Models

Cancer is a heterogeneous disease in that tumors of the same histology type can respond differently to a treatment. Anti-cancer drug response prediction is of paramount importance for both drug development and patient treatment design. Although various computational methods and data have been used to develop drug response prediction models, it remains a challenging problem due to the complexities of cancer mechanisms and cancer-drug interactions. To better characterize the interaction between cancer and drugs, we investigate the feasibility of integrating computationally derived features of molecular mechanisms of action into prediction models. Specifically, we add docking scores of drug molecules and target proteins in combination with cancer gene expressions and molecular drug descriptors for building response models. The results demonstrate a marginal improvement in drug response prediction performance when adding docking scores as additional features, through tests on large drug screening data. We discuss the limitations of the current approach and provide the research community with a baseline dataset of the large-scale computational docking for anti-cancer drugs.

Shape-Based Virtual Screening of a Billion-Compound Library Identifies Mycobacterial Lipoamide Dehydrogenase Inhibitors

Lpd (lipoamide dehydrogenase) in Mycobacterium tuberculosis (Mtb) is required for virulence and is a genetically validated tuberculosis (TB) target. Numerous screens have been performed over the last decade, yet only two inhibitor series have been identified. Recent advances in large-scale virtual screening methods combined with make-on-demand compound libraries have shown the potential for finding novel hits. In this study, the Enamine REAL library consisting of ∼1.12 billion compounds was efficiently screened using the GPU Shape screen method against Mtb Lpd to find additional chemical matter that would expand on the known sulfonamide inhibitor series. We identified six new inhibitors with IC50 in the range of 5-100 μM. While these compounds remained chemically close to the already known sulfonamide series inhibitors, some diversity was found in the cores of the hits. The two most potent hits were further validated by one-step potency optimization to submicromolar levels. The co-crystal structure of optimized analogue TDI-13537 provided new insights into the potency determinants of the series.

Structural and thermodynamic framework for PIEZO1 modulation by small molecules

Mechanosensitive PIEZO channels constitute potential pharmacological targets for multiple clinical conditions, spurring the search for potent chemical PIEZO modulators. Among them is Yoda1, a widely used synthetic small molecule PIEZO1 activator discovered through cell-based high-throughput screening. Yoda1 is thought to bind to PIEZO1's mechanosensory arm domain, sandwiched between two transmembrane regions near the channel pore. However, how the binding of Yoda1 to this region promotes channel activation remains elusive. Here, we first demonstrate that cross-linking PIEZO1 repeats A and B with disulfide bridges reduces the effects of Yoda1 in a redox-dependent manner, suggesting that Yoda1 acts by perturbing the contact between these repeats. Using molecular dynamics-based absolute binding free energy simulations, we next show that Yoda1 preferentially occupies a deeper, amphipathic binding site with higher affinity in PIEZO1 open state. Using Yoda1's binding poses in open and closed states, relative binding free energy simulations were conducted in the membrane environment, recapitulating structure-activity relationships of known Yoda1 analogs. Through virtual screening of an 8 million-compound library using computed fragment maps of the Yoda1 binding site, we subsequently identified two chemical scaffolds with agonist activity toward PIEZO1. This study supports a pharmacological model in which Yoda1 activates PIEZO1 by wedging repeats A and B, providing a structural and thermodynamic framework for the rational design of PIEZO1 modulators. Beyond PIEZO channels, the three orthogonal computational approaches employed here represent a promising path toward drug discovery in highly heterogeneous membrane protein systems.

Structure-based discovery of novel P-glycoprotein inhibitors targeting the nucleotide binding domains

P-glycoprotein (P-gp), a membrane transport protein overexpressed in certain drug-resistant cancer cells, has been the target of numerous drug discovery projects aimed at overcoming drug resistance in cancer. Most characterized P-gp inhibitors bind at the large hydrophobic drug binding domain (DBD), but none have yet attained regulatory approval. In this study, we explored the potential of designing inhibitors that target the nucleotide binding domains (NBDs), by computationally screening a large library of 2.6 billion synthesizable molecules, using a combination of machine learning-guided molecular docking and molecular dynamics (MD). 14 of the computationally best-scoring molecules were subsequently tested for their ability to inhibit P-gp mediated calcein-AM efflux. In total, five diverse compounds exhibited inhibitory effects in the calcein-AM assay without displaying toxicity. The activity of these compounds was confirmed by their ability to decrease the verapamil-stimulated ATPase activity of P-gp in a subsequent assay. The discovery of these five novel P-gp inhibitors demonstrates the potential of in-silico screening in drug discovery and provides a new stepping point towards future potent P-gp inhibitors.

Deep Learning with Geometry-Enhanced Molecular Representation for Augmentation of Large-Scale Docking-Based Virtual Screening

Structure-based virtual screening has been a crucial tool in drug discovery for decades. However, as the chemical space expands, the existing structure-based virtual screening techniques based on molecular docking and scoring struggle to handle billion-entry ultralarge libraries due to the high computational cost. To address this challenge, people have resorted to machine learning techniques to enhance structure-based virtual screening for efficiently exploring the vast chemical space. In those cases, compounds are usually treated as sequential strings or two-dimensional topology graphs, limiting their ability to incorporate three-dimensional structural information for downstream tasks. We herein propose a novel deep learning protocol, GEM-Screen, which utilizes the geometry-enhanced molecular representation of the compounds docking to a specific target and is trained on docking scores of a small fraction of a library through an active learning strategy to approximate the docking outcome for yet nontraining entries. This protocol is applied to virtual screening campaigns against the AmpC and D4 targets, demonstrating that GEM-Screen enriches more than 90% of the hit scaffolds for AmpC in the top 4% of model predictions and more than 80% of the hit scaffolds for D4 in the same top-ranking size of library. GEM-Screen can be used in conjunction with traditional docking programs for docking of only the top-ranked compounds to avoid the exhaustive docking of the whole library, thus allowing for discovering top-scoring compounds from billion-entry libraries in a rapid yet accurate fashion.

Molecular Dynamics for the Optimal Design of Functionalized Nanodevices to Target Folate Receptors on Tumor Cells

Atomistic details on the mechanism of targeting activity by biomedical nanodevices of specific receptors are still scarce in the literature, where mostly ligand/receptor pairs are modeled. Here, we use atomistic molecular dynamics (MD) simulations, free energy calculations, and machine learning approaches on the case study of spherical TiO2 nanoparticles (NPs) functionalized with folic acid (FA) as the targeting ligand of the folate receptor (FR). We consider different FA densities on the surface and different anchoring approaches, i.e., direct covalent bonding of FA γ-carboxylate or through polyethylene glycol spacers. By molecular docking, we first identify the lowest energy conformation of one FA inside the FR binding pocket from the X-ray crystal structure, which becomes the starting point of classical MD simulations in a realistic physiological environment. We estimate the binding free energy to be compared with the existing experimental data. Then, we increase complexity and go from the isolated FA to a nanosystem decorated with several FAs. Within the simulation time framework, we confirm the stability of the ligand-receptor interaction, even in the presence of the NP (with or without a spacer), and no significant modification of the protein secondary structure is observed. Our study highlights the crucial role played by the spacer, FA protonation state, and density, which are parameters that can be controlled during the nanodevice preparation step.

Docking for molecules that bind in a symmetric stack with SymDOCK

Discovering ligands for amyloid fibrils, such as those formed by the tau protein, is an area of much current interest. In recent structures, ligands bind in stacks in the tau fibrils to reflect the rotational and translational symmetry of the fibril itself; in these structures the ligands make few interactions with the protein but interact extensively with each other. To exploit this symmetry and stacking, we developed SymDOCK, a method to dock molecules that follow the protein's symmetry. For each prospective ligand pose, we apply the symmetry operation of the fibril to generate a self-interacting and fibril-interacting stack, checking that doing so will not cause a clash between the original molecule and its image. Absent a clash, we retain that pose and add the ligand-ligand van der Waals energy to the ligand's docking score (here using DOCK3.8). We can check these geometries and energies using an implementation of ANI, a neural network-based quantum-mechanical evaluation of the ligand stacking energies. In retrospective calculations, symmetry docking can reproduce the poses of three tau PET tracers whose structures have been determined. More convincingly, in a prospective study SymDOCK predicted the structure of the PET tracer MK-6240 bound in a symmetrical stack to AD PHF tau before that structure was determined; the docked pose was used to determine how MK-6240 fit the cryo-EM density. In proof-of-concept studies, SymDOCK enriched known ligands over property-matched decoys in retrospective screens without sacrificing docking speed, and can address large library screens that seek new symmetrical stackers. Future applications of this approach will be considered.

Precision Oncology Comes of Age: Designing Best-in-Class Small Molecules by Integrating Two Decades of Advances in Chemistry, Target Biology, and Data Science

Small-molecule drugs have enabled the practice of precision oncology for genetically defined patient populations since the first approval of imatinib in 2001. Scientific and technology advances over this 20-year period have driven the evolution of cancer biology, medicinal chemistry, and data science. Collectively, these advances provide tools to more consistently design best-in-class small-molecule drugs against known, previously undruggable, and novel cancer targets. The integration of these tools and their customization in the hands of skilled drug hunters will be necessary to enable the discovery of transformational therapies for patients across a wider spectrum of cancers.

SIGNIFICANCE

Target-centric small-molecule drug discovery necessitates the consideration of multiple approaches to identify chemical matter that can be optimized into drug candidates. To do this successfully and consistently, drug hunters require a comprehensive toolbox to avoid following the "law of instrument" or Maslow's hammer concept where only one tool is applied regardless of the requirements of the task. Combining our ever-increasing understanding of cancer and cancer targets with the technological advances in drug discovery described below will accelerate the next generation of small-molecule drugs in oncology.

Machine Learning-Boosted Docking Enables the Efficient Structure-Based Virtual Screening of Giga-Scale Enumerated Chemical Libraries

The emergence of ultra-large screening libraries, filled to the brim with billions of readily available compounds, poses a growing challenge for docking-based virtual screening. Machine learning (ML)-boosted strategies like the tool HASTEN combine rapid ML prediction with the brute-force docking of small fractions of such libraries to increase screening throughput and take on giga-scale libraries. In our case study of an anti-bacterial chaperone and an anti-viral kinase, we first generated a brute-force docking baseline for 1.56 billion compounds in the Enamine REAL lead-like library with the fast Glide high-throughput virtual screening protocol. With HASTEN, we observed robust recall of 90% of the true 1000 top-scoring virtual hits in both targets when docking only 1% of the entire library. This reduction of the required docking experiments by 99% significantly shortens the screening time. In the kinase target, the employment of a hydrogen bonding constraint resulted in a major proportion of unsuccessful docking attempts and hampered ML predictions. We demonstrate the optimization potential in the treatment of failed compounds when performing ML-boosted screening and benchmark and showcase HASTEN as a fast and robust tool in a growing arsenal of approaches to unlock the chemical space covered by giga-scale screening libraries for everyday drug discovery campaigns.

Recent Developments in Ultralarge and Structure-Based Virtual Screening Approaches

Drug development is a wide scientific field that faces many challenges these days. Among them are extremely high development costs, long development times, and a small number of new drugs that are approved each year. New and innovative technologies are needed to solve these problems that make the drug discovery process of small molecules more time and cost efficient, and that allow previously undruggable receptor classes to be targeted, such as protein-protein interactions. Structure-based virtual screenings (SBVSs) have become a leading contender in this context. In this review, we give an introduction to the foundations of SBVSs and survey their progress in the past few years with a focus on ultralarge virtual screenings (ULVSs). We outline key principles of SBVSs, recent success stories, new screening techniques, available deep learning-based docking methods, and promising future research directions. ULVSs have an enormous potential for the development of new small-molecule drugs and are already starting to transform early-stage drug discovery.

Profiling Chemobiological Connection between Natural Product and Target Space Based on Systematic Analysis

Natural products provide valuable starting points for new drugs with unique chemical structures. Here, we retrieve and join the LOTUS natural product database and ChEMBL interaction database to explore the relations and rhythm between chemical features of natural products and biotarget spaces. Our analysis revealed relations between the biogenic pathways of natural products and species taxonomy. Nitrogen-containing natural products were more likely to achieve high activity and have a higher potential to become candidate compounds. An apparent trend existed in the target space of natural products originating from different biological sources. Highly active alkaloids were more related to targets of neurodegenerative or neural diseases. Oligopeptides and polyketides were mainly associated with protein phosphorylation and HDAC receptors. Fatty acids readily intervened in various physiological processes involving prostanoids and leukotrienes. We also used FusionDTA, a deep learning model, to predict the affinity between all LOTUS natural products and 622 therapeutic drug targets, exploring the potential target space for natural products. Our data exploration provided a global perspective on the gaps in the chemobiological space of natural compounds through systematic analysis and prediction of their target space, which can be used for new drug design or natural drug repurposing.

A Simple Way to Incorporate Target Structural Information in Molecular Generative Models

Deep learning generative models are now being applied in various fields including drug discovery. In this work, we propose a novel approach to include target 3D structural information in molecular generative models for structure-based drug design. The method combines a message-passing neural network model that predicts docking scores with a generative neural network model as its reward function to navigate the chemical space searching for molecules that bind favorably with a specific target. A key feature of the method is the construction of target-specific molecular sets for training, designed to overcome potential transferability issues of surrogate docking models through a two-round training process. Consequently, this enables accurate guided exploration of the chemical space without reliance on the collection of prior knowledge about active and inactive compounds for the specific target. Tests on eight target proteins showed a 100-fold increase in hit generation compared to conventional docking calculations and the ability to generate molecules similar to approved drugs or known active ligands for specific targets without prior knowledge. This method provides a general and highly efficient solution for structure-based molecular generation.

Enhancing Hit Discovery in Virtual Screening through Absolute Protein-Ligand Binding Free-Energy Calculations

In the hit identification stage of drug discovery, a diverse chemical space needs to be explored to identify initial hits. Contrary to empirical scoring functions, absolute protein-ligand binding free-energy perturbation (ABFEP) provides a theoretically more rigorous and accurate description of protein-ligand binding thermodynamics and could, in principle, greatly improve the hit rates in virtual screening. In this work, we describe an implementation of an accurate and reliable ABFEP method in FEP+. We validated the ABFEP method on eight congeneric compound series binding to eight protein receptors including both neutral and charged ligands. For ligands with net charges, the alchemical ion approach is adopted to avoid artifacts in electrostatic potential energy calculations. The calculated binding free energies correlate with experimental results with a weighted average of R² = 0.55 for the entire dataset. We also observe an overall root-mean-square error (RMSE) of 1.1 kcal/mol after shifting the zero-point of the simulation data to match the average experimental values. Through ABFEP calculations using apo versus holo protein structures, we demonstrated that the protein conformational and protonation state changes between the apo and holo proteins are the main physical factors contributing to the protein reorganization free energy manifested by the overestimation of raw ABFEP calculated binding free energies using the holo structures of the proteins. Furthermore, we performed ABFEP calculations in three virtual screening applications for hit enrichment. ABFEP greatly improves the hit rates as compared to docking scores or other methods like metadynamics. The good performance of ABFEP in rank ordering compounds demonstrated in this work confirms it as a useful tool to improve the hit rates in virtual screening, thus facilitating hit discovery.

Uni-Dock: GPU-Accelerated Docking Enables Ultralarge Virtual Screening

Molecular docking, a structure-based virtual screening method, is a reliable tool to enrich potential bioactive molecules from molecular databases. With the rapid expansion of compound library sizes, the speed of existing molecular docking programs becomes less than adequate to meet the demand for screening ultralarge libraries containing tens of millions or billions of molecules. Here, we propose Uni-Dock, a GPU-accelerated molecular docking program that supports various scoring functions including vina, vinardo, and ad4. Uni-Dock achieves more than 1000-fold speedup with high accuracy compared with the AutoDock Vina running in single CPU core, outperforming reported GPU-accelerated docking programs including AutoDock-GPU and Vina-GPU based on head-to-head experiments. Uni-Dock docks molecules in batches simultaneously using concurrent threads of each molecule. The data flow between GPU and CPU is optimized to eliminate CPU hotspots and maximize GPU utility. Additionally, Uni-Dock also supports hydrogen bond biased docking for all scoring functions and can be migrated to multiple GPUs of different architectures and manufacturers. We analyzed the improved performance of Uni-Dock on the CASF-2016 and DUD-E datasets and recommend three combinations of hyperparameters corresponding to different docking scenarios. To demonstrate Uni-Dock's capability on routinely screening ultralarge libraries, we performed hierarchical virtual screening experiments with Uni-Dock on the Enamine Diverse REAL druglike set containing 38.2 million molecules to a popular target KRAS G12D in 12 h using 100 NVIDIA V100 GPUs. To the best of our knowledge, Uni-Dock should be the fastest GPU-accelerated docking program to date.

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.

Epik: pKa and Protonation State Prediction through Machine Learning

Epik version 7 is a software program that uses machine learning for predicting the pK_a values and protonation state distribution of complex, druglike molecules. Using an ensemble of atomic graph convolutional neural networks (GCNNs) trained on over 42,000 pK_a values across broad chemical space from both experimental and computed origins, the model predicts pK_a values with 0.42 and 0.72 pK_a unit median absolute and root mean square errors, respectively, across seven test sets. Epik version 7 also generates protonation states and recovers 95% of the most populated protonation states compared to previous versions. Requiring on average only 47 ms per ligand, Epik version 7 is rapid and accurate enough to evaluate protonation states for crucial molecules and prepare ultra-large libraries of compounds to explore vast regions of chemical space. The simplicity and time required for the training allow for the generation of highly accurate models customized to a program's specific chemistry.

Computational approaches streamlining drug discovery

Computer-aided drug discovery has been around for decades, although the past few years have seen a tectonic shift towards embracing computational technologies in both academia and pharma. This shift is largely defined by the flood of data on ligand properties and binding to therapeutic targets and their 3D structures, abundant computing capacities and the advent of on-demand virtual libraries of drug-like small molecules in their billions. Taking full advantage of these resources requires fast computational methods for effective ligand screening. This includes structure-based virtual screening of gigascale chemical spaces, further facilitated by fast iterative screening approaches. Highly synergistic are developments in deep learning predictions of ligand properties and target activities in lieu of receptor structure. Here we review recent advances in ligand discovery technologies, their potential for reshaping the whole process of drug discovery and development, as well as the challenges they encounter. We also discuss how the rapid identification of highly diverse, potent, target-selective and drug-like ligands to protein targets can democratize the drug discovery process, presenting new opportunities for the cost-effective development of safer and more effective small-molecule treatments.

Industrializing AI/ML during the end-to-end drug discovery process

Drug discovery aims to select proper targets and drug candidates to address unmet clinical needs. The end-to-end drug discovery process includes all stages of drug discovery from target identification to drug candidate selection. Recently, several artificial intelligence and machine learning (AI/ML)-based drug discovery companies have attempted to build data-driven platforms spanning the end-to-end drug discovery process. The ability to identify elusive targets essentially leads to the diversification of discovery pipelines, thereby increasing the ability to address unmet needs. Modern ML technologies are complementing traditional computer-aided drug discovery by accelerating candidate optimization in innovative ways. This review summarizes recent developments in AI/ML methods from target identification to molecule optimization, and concludes with an overview of current industrial trends in end-to-end AI/ML platforms.

Chemical Space Virtual Screening against Hard-to-Drug RNA Methyltransferases DNMT2 and NSUN6

Targeting RNA methyltransferases with small molecules as inhibitors or tool compounds is an emerging field of interest in epitranscriptomics and medicinal chemistry. For two challenging RNA methyltransferases that introduce the 5-methylcytosine (m⁵C) modification in different tRNAs, namely DNMT2 and NSUN6, an ultra-large commercially available chemical space was virtually screened by physicochemical property filtering, molecular docking, and clustering to identify new ligands for those enzymes. Novel chemotypes binding to DNMT2 and NSUN6 with affinities down to K_D,app = 37 µM and K_D,app = 12 µM, respectively, were identified using a microscale thermophoresis (MST) binding assay. These compounds represent the first molecules with a distinct structure from the cofactor SAM and have the potential to be developed into activity-based probes for these enzymes. Additionally, the challenges and strategies of chemical space docking screens with special emphasis on library focusing and diversification are discussed.

From In Silico to a Cellular Model: Molecular Docking Approach to Evaluate Antioxidant Bioactive Peptides

The increasing need to counteract the redox imbalance in chronic diseases leads to focusing research on compounds with antioxidant activity. Among natural molecules with health-promoting effects on many body functions, bioactive peptides are gaining interest. They are protein fragments of 2–20 amino acids that can be released by various mechanisms, such as gastrointestinal digestion, food processing and microbial fermentation. Recent studies report the effects of bioactive peptides in the cellular environment, and there is evidence that these compounds can exert their action by modulating specific pathways. This review focuses on the newest approaches to the structure–function correlation of the antioxidant bioactive peptides, considering their molecular mechanism, by evaluating the activation of specific signaling pathways that are linked to antioxidant systems. The correlation between the results of in silico molecular docking analysis and the effects in a cellular model was highlighted. This knowledge is fundamental in order to propose the use of bioactive peptides as ingredients in functional foods or nutraceuticals.

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.

ZINC-22─A Free Multi-Billion-Scale Database of Tangible Compounds for Ligand Discovery

Purchasable chemical space has grown rapidly into the tens of billions of molecules, providing unprecedented opportunities for ligand discovery but straining the tools that might exploit these molecules at scale. We have therefore developed ZINC-22, a database of commercially accessible small molecules derived from multi-billion-scale make-on-demand libraries. The new database and tools enable analog searching in this vast new space via a facile GUI, CartBlanche, drawing on similarity methods that scale sublinearly in the number of molecules. The new library also uses data organization methods, enabling rapid lookup of molecules and their physical properties, including conformations, partial atomic charges, c Log P values, and solvation energies, all crucial for molecule docking, which had become slow with older database organizations in previous versions of ZINC. As the libraries have continued to grow, we have been interested in finding whether molecular diversity has suffered, for instance, because certain scaffolds have come to dominate via easy analoging. This has not occurred thus far, and chemical diversity continues to grow with database size, with a log increase in Bemis-Murcko scaffolds for every two-log unit increase in database size. Most new scaffolds come from compounds with the highest heavy atom count. Finally, we consider the implications for databases like ZINC as the libraries grow toward and beyond the trillion-molecule range. ZINC is freely available to everyone and may be accessed at cartblanche22.docking.org, via Globus, and in the Amazon AWS and Oracle OCI clouds.

'Chemistry at the speed of sound': automated 1536-well nanoscale synthesis of 16 scaffolds in parallel

Screening of large and diverse libraries is the 'bread and butter' in the first phase of the discovery of novel drugs. However, maintenance and periodic renewal of high-quality large compound collections pose considerable logistic, environmental and monetary problems. Here, we exercise an alternative, the 'on-the-fly' synthesis of large and diverse libraries on a nanoscale in a highly automated fashion. For the first time, we show the feasibility of the synthesis of a large library based on 16 different chemistries in parallel on several 384-well plates using the acoustic dispensing ejection (ADE) technology platform. In contrast to combinatorial chemistry, we produced 16 scaffolds at the same time and in a sparse matrix fashion, and each compound was produced by a random combination of diverse large building blocks. The synthesis, analytics, resynthesis of selected compounds, and chemoinformatic analysis of the library are described. The advantages of the herein described automated nanoscale synthesis approach include great library diversity, absence of library storage logistics, superior economics, speed of synthesis by automation, increased safety, and hence sustainable chemistry.

Benzoquinoline Chemical Space: A Helpful Approach in Antibacterial and Anticancer Drug Design

Benzoquinolines are used in many drug design projects as starting molecules subject to derivatization. This computational study aims to characterize e benzoquinone drug space to ease future drug design processes based on these molecules. The drug space is composed of all benzoquinones, which are active on topoisomerase II and ATP synthase. Topological, chemical, and bioactivity spaces are explored using computational methodologies based on virtual screening and scaffold hopping and molecular docking, respectively. Topological space is a geometrical space in which the elements composing it can be defined as a set of neighbors (which satisfy a particular axiom). In such space, a chemical space can be defined as the property space spanned by all possible molecules and chemical compounds adhering to a given set of construction principles and boundary conditions. In this chemical space, the potentially pharmacologically active molecules form the bioactivity space. Results show a poly-morphological chemical space that suggests distinct characteristics. The chemical space is correlated with properties such as steric energy, the number of hydrogen bonds, the presence of halogen atoms, and membrane permeability-related properties. Lastly, novel chemical compounds (such as oxadiazole methybenzamide and floro methylcyclohexane diene) with drug-like potential, active on TOPO II and ATP synthase have been identified.

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.

Efficient design of peptide-binding polymers using active learning approaches

Active learning (AL) has become a subject of active recent research both in industry and academia as an efficient approach for rapid design and discovery of novel chemicals, materials, and polymers. Herein, we have assessed the applicability of AL for the discovery of polymeric micelle formulations for poorly soluble drugs. We were motivated by the key advantages of this approach making it a desirable strategy for rational design of drug delivery systems due toto its ability to (i) employ relatively small datasets for model development, (ii) iterate between model development and model assessment using small external datasets that can be either generated in focused experimental studies or formed from subsets of the initial training data, and (iii) progressively evolve models towards increasingly more reliable predictions and the identification of novel chemicals with the desired properties. In this study, we compared various AL protocols for their effectiveness in finding biologically active molecules using synthetic datasets. We have investigated the dependency of AL performance on the size of the initial training set, the relative complexity of the task, and the choice of the initial training dataset. We found that AL techniques as applied to regression modeling offer no benefits over random search, while AL used for classification tasks performs better than models built for randomly selected training sets but still quite far from perfect. Using the best performing AL protocol,. Finally, the best performing AL approach was employed to discover and experimentally validate novel binding polymers for a case study of asialoglycoprotein receptor (ASGPR).

Magnet for the Needle in Haystack: "Crystal Structure First" Fragment Hits Unlock Active Chemical Matter Using Targeted Exploration of Vast Chemical Spaces

Fragment-based drug discovery (FBDD) has successfully led to approved therapeutics for challenging and "undruggable" targets. In the context of FBDD, we introduce a novel, multidisciplinary method to identify active molecules from purchasable chemical space. Starting from four small-molecule fragment complexes of protein kinase A (PKA), a template-based docking screen using Enamine's multibillion REAL Space was performed. A total of 93 molecules out of 106 selected compounds were successfully synthesized. Forty compounds were active in at least one validation assay with the most active follow-up having a 13,500-fold gain in affinity. Crystal structures for six of the most promising binders were rapidly obtained, verifying the binding mode. The overall success rate for this novel fragment-to-hit approach was 40%, accomplished in only 9 weeks. The results challenge the established fragment prescreening paradigm since the standard industrial filters for fragment hit identification in a thermal shift assay would have missed the initial fragments.

Chemical Space Exploration with Active Learning and Alchemical Free Energies

Drug discovery can be thought of as a search for a needle in a haystack: searching through a large chemical space for the most active compounds. Computational techniques can narrow the search space for experimental follow up, but even they become unaffordable when evaluating large numbers of molecules. Therefore, machine learning (ML) strategies are being developed as computationally cheaper complementary techniques for navigating and triaging large chemical libraries. Here, we explore how an active learning protocol can be combined with first-principles based alchemical free energy calculations to identify high affinity phosphodiesterase 2 (PDE2) inhibitors. We first calibrate the procedure using a set of experimentally characterized PDE2 binders. The optimized protocol is then used prospectively on a large chemical library to navigate toward potent inhibitors. In the active learning cycle, at every iteration a small fraction of compounds is probed by alchemical calculations and the obtained affinities are used to train ML models. With successive rounds, high affinity binders are identified by explicitly evaluating only a small subset of compounds in a large chemical library, thus providing an efficient protocol that robustly identifies a large fraction of true positives.

Batched Bayesian Optimization for Drug Design in Noisy Environments

The early stages of the drug design process involve identifying compounds with suitable bioactivities via noisy assays. As databases of possible drugs are often very large, assays can only be performed on a subset of the candidates. Selecting which assays to perform is best done within an active learning process, such as batched Bayesian optimization, and aims to reduce the number of assays that must be performed. We compare how noise affects different batched Bayesian optimization techniques and introduce a retest policy to mitigate the effect of noise. Our experiments show that batched Bayesian optimization remains effective, even when large amounts of noise are present, and that the retest policy enables more active compounds to be identified in the same number of experiments.

Uncertainty quantification: Can we trust artificial intelligence in drug discovery?

The problem of human trust is one of the most fundamental problems in applied artificial intelligence in drug discovery. In silico models have been widely used to accelerate the process of drug discovery in recent years. However, most of these models can only give reliable predictions within a limited chemical space that the training set covers (applicability domain). Predictions of samples falling outside the applicability domain are unreliable and sometimes dangerous for the drug-design decision-making process. Uncertainty quantification accordingly has drawn great attention to enable autonomous drug designing. By quantifying the confidence level of model predictions, the reliability of the predictions can be quantitatively represented to assist researchers in their molecular reasoning and experimental design. Here we summarize the state-of-the-art approaches to uncertainty quantification and underline how they can be used for drug design and discovery projects. Furthermore, we also outline four representative application scenarios of uncertainty quantification in drug discovery.

Parallel tempered genetic algorithm guided by deep neural networks for inverse molecular design

Inverse molecular design involves algorithms that sample molecules with specific target properties from a multitude of candidates and can be posed as an optimization problem. High-dimensional optimization tasks in the natural sciences are commonly tackled via population-based metaheuristic optimization algorithms such as evolutionary algorithms. However, often unavoidable expensive property evaluation can limit the widespread use of such approaches as the associated cost can become prohibitive. Herein, we present JANUS, a genetic algorithm inspired by parallel tempering. It propagates two populations, one for exploration and another for exploitation, improving optimization by reducing property evaluations. JANUS is augmented by a deep neural network that approximates molecular properties and relies on active learning for enhanced molecular sampling. It uses the SELFIES representation and the STONED algorithm for the efficient generation of structures, and outperforms other generative models in common inverse molecular design tasks achieving state-of-the-art target metrics across multiple benchmarks. As neither most of the benchmarks nor the structure generator in JANUS account for synthesizability, a significant fraction of the proposed molecules is synthetically infeasible demonstrating that this aspect needs to be considered when evaluating the performance of molecular generative models.

Virtual Screening in the Cloud Identifies Potent and Selective ROS1 Kinase Inhibitors

ROS1 rearrangements account for 1-2% of non-small cell lung cancer patients, yet there are no specifically designed, selective ROS1 therapies in the clinic. Previous knowledge of potent ROS1 inhibitors with selectivity over TrkA, a selected antitarget, enabled virtual screening as a hit finding approach in this project. The ligand-based virtual screening was focused on identifying molecules with a similar 3D shape and pharmacophore to the known actives. To that end, we turned to the AstraZeneca virtual library, estimated to cover 10¹⁵ synthesizable make-on-demand molecules. We used cloud computing-enabled FastROCS technology to search the enumerated 10¹⁰ subset of the full virtual space. A small number of specific libraries were prioritized based on the compound properties and a medicinal chemistry assessment and further enumerated with available building blocks. Following the docking evaluation to the ROS1 structure, the most promising hits were synthesized and tested, resulting in the identification of several potent and selective series. The best among them gave a nanomolar ROS1 inhibitor with over 1000-fold selectivity over TrkA and, from the preliminary established SAR, these have the potential to be further optimized. Our prospective study describes how conceptually simple shape-matching approaches can identify potent and selective compounds by searching ultralarge virtual libraries, demonstrating the applicability of such workflows and their importance in early drug discovery.

Exploration of Ultralarge Compound Collections for Drug Discovery

Designing new medicines more cheaply and quickly is tightly linked to the quest of exploring chemical space more widely and efficiently. Chemical space is monumentally large, but recent advances in computer software and hardware have enabled researchers to navigate virtual chemical spaces containing billions of chemical structures. This review specifically concerns collections of many millions or even billions of enumerated chemical structures as well as even larger chemical spaces that are not fully enumerated. We present examples of chemical libraries and spaces and the means used to construct them, and we discuss new technologies for searching huge libraries and for searching combinatorially in chemical space. We also cover space navigation techniques and consider new approaches to de novo drug design and the impact of the "autonomous laboratory" on synthesis of designed compounds. Finally, we summarize some other challenges and opportunities for the future.

Systemic evolutionary chemical space exploration for drug discovery

Chemical space exploration is a major task of the hit-finding process during the pursuit of novel chemical entities. Compared with other screening technologies, computational de novo design has become a popular approach to overcome the limitation of current chemical libraries. Here, we reported a de novo design platform named systemic evolutionary chemical space explorer (SECSE). The platform was conceptually inspired by fragment-based drug design, that miniaturized a “lego-building” process within the pocket of a certain target. The key to virtual hits generation was then turned into a computational search problem. To enhance search and optimization, human intelligence and deep learning were integrated. Application of SECSE against phosphoglycerate dehydrogenase (PHGDH), proved its potential in finding novel and diverse small molecules that are attractive starting points for further validation. This platform is open-sourced and the code is available at http://github.com/KeenThera/SECSE.

Discovery of Kinase and Carbonic Anhydrase Dual Inhibitors by Machine Learning Classification and Experiments

A multi-target small molecule modulator is advantageous for treating complicated diseases such as cancers. However, the strategy and application for discovering a multi-target modulator have been less reported. This study presents the dual inhibitors for kinase and carbonic anhydrase (CA) predicted by machine learning (ML) classifiers, and validated by biochemical and biophysical experiments. ML trained by CA I and CA II inhibitor molecular fingerprints predicted candidates from the protein-specific bioactive molecules approved or under clinical trials. For experimental tests, three sulfonamide-containing kinase inhibitors, 5932, 5946, and 6046, were chosen. The enzyme assays with CA I, CA II, CA IX, and CA XII have allowed the quantitative comparison in the molecules’ inhibitory activities. While 6046 inhibited weakly, 5932 and 5946 exhibited potent inhibitions with 100 nM to 1 μM inhibitory constants. The ML screening was extended for finding CAs inhibitors of all known kinase inhibitors. It found XMU-MP-1 as another potent CA inhibitor with an approximate 30 nM inhibitory constant for CA I, CA II, and CA IX. Differential scanning fluorimetry confirmed the direct interaction between CAs and small molecules. Cheminformatics studies, including docking simulation, suggest that each molecule possesses two separate functional moieties: one for interaction with kinases and the other with CAs.

A highly accurate metadynamics-based Dissociation Free Energy method to calculate protein-protein and protein-ligand binding potencies

Although seeking to develop a general and accurate binding free energy calculation method for protein-protein and protein-ligand interactions has been a continuous effort for decades, only limited successes have been obtained so far. Here, we report the development of a metadynamics-based procedure that calculates Dissociation Free Energy (DFE) and its application to 19 non-congeneric protein-protein complexes and hundreds of protein-ligand complexes covering eight targets. We achieved very high correlations in comparison to experimental binding free energies for these diverse sets of systems, demonstrating the generality and accuracy of the method. Since structures of most proteins are available owing to the recent success of prediction by artificial intelligence, a general free energy method such as DFE, combined with other methods, can make structure-based drug design a widely viable and reliable solution to develop both traditional small molecule drugs and biologic drugs as well as PROTACS.

Integration of Molecular Docking Analysis and Molecular Dynamics Simulations for Studying Food Proteins and Bioactive Peptides

In silico tools, such as molecular docking, are widely applied to study interactions and binding affinity of biological activity of proteins and peptides. However, restricted sampling of both ligand and receptor conformations and use of approximated scoring functions can produce results that do not correlate with actual experimental binding affinities. Molecular dynamics simulations (MDS) can provide valuable information in deciphering functional mechanisms of proteins/peptides and other biomolecules, overcoming the rigid sampling limitations in docking analysis. This review will discuss the information related to the traditional use of in silico models, such as molecular docking, and its application for studying food proteins and bioactive peptides, followed by an in-depth introduction to the theory of MDS and description of why these molecular simulation techniques are important in the theoretical prediction of structural and functional dynamics of food proteins and bioactive peptides. Applications, limitations, and future prospects of MDS will also be discussed.

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.

Automated discovery of noncovalent inhibitors of SARS-CoV-2 main protease by consensus Deep Docking of 40 billion small molecules

Recent explosive growth of 'make-on-demand' chemical libraries brought unprecedented opportunities but also significant challenges to the field of computer-aided drug discovery. To address this expansion of the accessible chemical universe, molecular docking needs to accurately rank billions of chemical structures, calling for the development of automated hit-selecting protocols to minimize human intervention and error. Herein, we report the development of an artificial intelligence-driven virtual screening pipeline that utilizes Deep Docking with Autodock GPU, Glide SP, FRED, ICM and QuickVina2 programs to screen 40 billion molecules against SARS-CoV-2 main protease (Mpro). This campaign returned a significant number of experimentally confirmed inhibitors of Mpro enzyme, and also enabled to benchmark the performance of twenty-eight hit-selecting strategies of various degrees of stringency and automation. These findings provide new starting scaffolds for hit-to-lead optimization campaigns against Mpro and encourage the development of fully automated end-to-end drug discovery protocols integrating machine learning and human expertise.

Enhancing preclinical drug discovery with artificial intelligence

Artificial intelligence (AI) is becoming an integral part of drug discovery. It has the potential to deliver across the drug discovery and development value chain, starting from target identification and reaching through clinical development. In this review, we provide an overview of current AI technologies and a glimpse of how AI is reimagining preclinical drug discovery by highlighting examples where AI has made a real impact. Considering the excitement and hyperbole surrounding AI in drug discovery, we aim to present a realistic view by discussing both opportunities and challenges in adopting AI in drug discovery.