Rules were made to be broken

Exploring the Potential of Adaptive, Local Machine Learning in Comparison to the Prediction Performance of Global Models: A Case Study from Bayer's Caco-2 Permeability Database

Machine learning (ML) techniques are being widely implemented to fill the gap in simple molecular design guidelines for newer therapeutic modalities in the extended and beyond rule of five chemical space (eRo5, bRo5). These ML techniques predict molecular properties directly from the structure, allowing for the prioritization of promising compounds. However, the performance of models varies greatly among ML use cases. A molecular property for which achieving sufficient performance in generalizing global models still remains difficult is Caco-2 permeability. Especially within the lower permeability ranges, which are specific for larger molecules belonging to the e/bRo5 space, accurate regression predictions have proven to be challenging. The present study, therefore, identifies a suitable combination of ML algorithm and descriptors, consisting of the LightGBM algorithm and RDKit molecular property descriptors, to predict Caco-2 permeability very efficiently by a simple global model. An additionally introduced local model uses the same algorithm and descriptors but selects its training data based on Tanimoto fingerprint similarity to match the individual test compound's structure. Evaluation of this adaptive model, by systematically varying the number of most similar structures for training, shows that, in comparison to the global model, there was only marginally improved performance with specific training data constellations. These random improvements indicate that deriving general rules for local model parametrization is not possible a priori for the chosen algorithm and descriptor combination, and preselecting training data does not seem advantageous over global ML based on all available data, while creation of more data-efficient models was generally proven to be possible.

Closing the Design-Make-Test-Analyze Loop: Interplay between Experiments and Predictions Drives PROTACs Bioavailability

The drug development landscape is expanding to include drug modalities such as PROteolysis-TArgeting Chimeras (PROTACs) and peptides, offering possibilities for previously intractable biological targets. However, with their size and chemical nature, they diverge from established frameworks for the prediction of oral bioavailability. This evolution to larger and more complex molecules necessitates new methodologies and prediction models to continuously expand on bioavailability guidelines. We describe the high-capacity adoption of two chromatographic physicochemical assays and their application for iterative compound optimization to achieve oral bioavailability. We further describe how these data underpin the continuous refinement of internal machine learning models, which guide compound synthesis decisions in the molecular design phase. Based on data for a set of 691 PROTACs, and two project examples, we confirm a sweet spot for oral bioavailability at log D values higher than the norm for small molecules and show how experimental data and prediction models synergize to effectively drive chemistry optimization.

How Should we Teach Medicinal Chemistry in Higher Education to Prepare Students for a Future Career as Medicinal Chemists and Drug Designers? - A Teacher's Perspective

In the recent two decades, the multidisciplinary field of medicinal chemistry has undergone several conceptual and technology-driven paradigm changes with significant impact on the skill set medicinal chemists need to acquire during their education. Considering the need for academic medicinal chemistry teaching, this article aims at identifying important skills, competences, and basic knowledge as general learning outcomes based on an analysis of the relevant stakeholders and concludes effective teaching strategies preparing students for a future career as medicinal chemists and drug designers.

Property-based optimisation of PROTACs

PROTACs are an emerging therapeutic approach towards targeted protein degradation. This article examines the leading examples of this modality that are in clinical development through the prism of their physicochemical properties. In particular, the optimisation of the various components of PROTACs together with the difficulties faced by medicinal chemists seeking to achieve oral bioavailability in this challenging space are outlined. Guidance, opinion and advice based on the authors' own experiences in this area are offered in the hope this may be useful to others working in this fascinating frontier of drug discovery.

A computational assay for identifying millet-derived compounds that antagonize the interaction between bisphenols and estrogen-related receptor gamma

The use of Bisphenol A (BPA) and its analogs in industries, as well as the products made from them, is becoming a significant concern for human health. Scientific studies have revealed that BPA functions as an endocrine disruptor. While some analogs of BPA (bisphenols) have been used for a longer time, it was later discovered that they are toxic, similar to BPA. Their widespread use ensures their presence in the environment, and thus, everyone is exposed to them. Scientific research has shown that BPA interacts with estrogen-related receptor gamma (ERRγ), affecting its normal function. ERRγ is involved in biological processes including energy metabolism and mitochondrial function. Therefore, continuous exposure to bisphenols increases the risk of various diseases. In our previous study, we observed that some analogs of BPA had a higher binding affinity to ERRγ compared to BPA itself and analyzed the amino acid residues involved in this interaction. We hypothesized that by antagonizing the interaction between bisphenols and ERRγ, we could neutralize their toxic effects. Taking into account the health benefits of millets and their toxin removal properties, virtual screening of millet-derived compounds was conducted along with prediction of their ADMET profiles. Top five candidates were prioritized for Density Functional Theory (DFT) calculations and further analyses. Long-term molecular dynamics simulation (1 µs) were utilized to evaluate their binding, stability, and antagonizing abilities. Furthermore, reevaluation of their binding energy was conducted using the MM-PBSA method. This study reports millet-derived compounds, namely, Tricin 7-rutinoside, Tricin 7-glucoside, Glucotricin, Kaempferol, and Setarin. These compounds are predicted to be potent competitive inhibitors that can antagonize the interactions between bisphenols and ERRγ. These compounds could potentially assist in the development of future therapeutics. They may also be considered for use as food supplements, although further investigations, including wet-lab experiments and clinical studies, are needed.

Validating Small-Molecule Force Fields for Macrocyclic Compounds Using NMR Data in Different Solvents

Macrocycles are a promising class of compounds as therapeutics for difficult drug targets due to a favorable combination of properties: They often exhibit improved binding affinity compared to their linear counterparts due to their reduced conformational flexibility, while still being able to adapt to environments of different polarity. To assist in the rational design of macrocyclic drugs, there is need for computational methods that can accurately predict conformational ensembles of macrocycles in different environments. Molecular dynamics (MD) simulations remain one of the most accurate methods to predict ensembles quantitatively, although the accuracy is governed by the underlying force field. In this work, we benchmark four different force fields for their application to macrocycles by performing replica exchange with solute tempering (REST2) simulations of 11 macrocyclic compounds and comparing the obtained conformational ensembles to nuclear Overhauser effect (NOE) upper distance bounds from NMR experiments. Especially, the modern force fields OpenFF 2.0 and XFF yield good results, outperforming force fields like GAFF2 and OPLS/AA. We conclude that REST2 in combination with modern force fields can often produce accurate ensembles of macrocyclic compounds. However, we also highlight examples for which all examined force fields fail to produce ensembles that fulfill the experimental constraints.

Structure-based identification of small-molecule inhibitors that target the DIII domain of the Dengue virus glycoprotein E pan-serotypically

Dengue viral infection is caused by the Dengue virus, which spreads to humans through the bite of infected mosquitos. Dengue affects over half of the global population, with an estimated 500 million infections per year. Despite this, no effective treatment is currently available, however, several promising candidates are undergoing pre-clinical/clinical testing. The existence of four major serotypes is an important challenge in the development of drugs and vaccines to combat Dengue virus infection. Hence, the drug/vaccine thereby developed should neutralize all the four serotypes equally. However, there is no pan-serotype specific treatment for Dengue virus, thereby emphasizing the need for the identification of novel drug-like compounds that can target all serotypes of the Dengue virus equally. To this end, we employed virtual screening methodologies to find drug-like compounds that target the domain III of glycoprotein E. Most importantly, domain III of E protein is involved in viral fusion with the host membrane and is also targeted by neutralizing antibodies. Our study found two small molecule drug-like compounds (out of the 3 million compounds screened) having similar binding affinity with all four serotypes. The compounds thereby identified exhibit favourable drug like properties and can be developed as a treatment for Dengue virus.

The IMS Library: from IN-Stock to Virtual

A chemical library is a key element in the early stages of pharmaceutical research. Its design encompasses various factors, such as diversity, size, ease of synthesis, aimed at increasing the likelihood of success in drug discovery. This article explores the collaborative efforts of computational and synthetic chemists in tailoring chemical libraries for cost-effective and resource-efficient use, particularly in the context of academic research projects. It proposes chemoinformatics methodologies that address two pivotal questions: first, crafting a diverse panel of under 1000 compounds from an existing pool through synthetic efforts, leveraging the expertise of organic chemists; and second, expanding pharmacophoric diversity within this panel by creating a highly accessible virtual chemical library. Chemoinformatics tools were developed to analyse initial panel of about 10,000 compounds into two tailored libraries: eIMS and vIMS. The eIMS Library comprises 578 diverse in-stock compounds ready for screening. Its virtual counterpart, vIMS, features novel compounds guided by chemists, ensuring synthetic accessibility. vIMS offers a broader array of binding motifs and improved drug-like characteristics achieved through the addition of diverse functional groups to eIMS scaffolds followed by filtering of reactive or unusual structures. The uniqueness of vIMS is emphasized through a comparison with commercial suppliers' virtual chemical space.

Unlocking the potential of higher-molecular-weight 5-HT7R ligands: Synthesis, affinity, and ADMET examination

An increasing number of drugs introduced to the market and numerous repositories of compounds with confirmed activity have posed the need to revalidate the state-of-the-art rules that determine the ranges of properties the compounds should possess to become future drugs. In this study, we designed a series of two chemotypes of aryl-piperazine hydantoin ligands of 5-HT7R, an attractive target in search for innovative CNS drugs, with higher molecular weight (close to or over 500). Consequently, 14 new compounds were synthesised and screened for their receptor activity accompanied by extensive docking studies to evaluate the observed structure-activity/properties relationships. The ADMET characterisation in terms of the biological membrane permeability, metabolic stability, hepatotoxicity, cardiotoxicity, and protein plasma binding of the obtained compounds was carried out in vitro. The outcome of these studies constituted the basis for the comprehensive challenge of computational tools for ADMET properties prediction. All the compounds possessed high affinity to the 5-HT7R (Ki below 250 nM for all analysed structures) with good selectivity over 5-HT6R and varying affinity towards 5-HT2AR, 5-HT1AR and D2R. For the best compounds of this study, the expression profile of genes associated with neurodegeneration, anti-oxidant response and anti-inflammatory function was determined, and the survival of the cells (SH-SY5Y as an in vitro model of Alzheimer's disease) was evaluated. One 5-HT7R agent (32) was characterised by a very promising ADMET profile, i.e. good membrane permeability, low hepatotoxicity and cardiotoxicity, and high metabolic stability with the simultaneous high rate of plasma protein binding and high selectivity over other GPCRs considered, together with satisfying gene expression profile modulations and neural cell survival. Such encouraging properties make it a good candidate for further testing and optimisation as a potential agent in the treatment of CNS-related disorders.

Between Theory and Practice: Computational/Experimental Integrated Approaches to Understand the Solubility and Lipophilicity of PROTACs

Emerging drug candidates more often fall in the beyond-rule-of-five chemical space. Among them, proteolysis targeting chimeras (PROTACs) have gained great attention in the past decade. Although physicochemical properties of small molecules accomplishing Lipinski's rule-of-five can now be easily predicted through models generated by large data collections, for PROTACs the knowledge is still limited and heterogeneous, hampering their prediction. Here, the kinetic solubility and the coefficient of distribution at pH 7.4 (LogD7.4) of 44 PROTACs, designed and synthesized to cover a wide chemical space, were measured. Their generally low solubility and high lipophilicity required an optimization of the experimental methods. Concerning the LogD7.4, several in silico prediction tools were tested, which were quite accurate for classical small molecules but provided dissimilar outcomes for PROTACs. Finally, in silico models for the prediction of PROTACs' kinetic solubility and LogD7.4 were proposed by combining in-house generated experimental data with 3D description of PROTACs' structures.

Immobilized cells on microcarriers for efficient and biomimetic screening of active compounds acting on FGFR4 from Fructus evodiae

Cell membrane coating strategies have been increasingly researched in new drug discovery from complex herb extracts. However, these systems failed to maintain the functionality of the coated cells because cell membranes, not whole cells were used. Original source cells can be used as a vector for active compound screening in a manner that mimics in vivo processes. In this study, we established a novel approach to fabricate high-density fibroblast growth factor receptor 4 (FGFR4)-HEK293 cells on microcarriers covered with collagen through cell culture and covalent immobilization between proteins. This method enables the efficient screening of active compounds from herbs. Two compounds, evodiamine and limonin, were obtained from Fructus evodiae, which were proven to inhibit the FGFR4 target. Enhanced immobilization effects and negligible damage to FGFR4-HEK293 cells treated with paraformaldehyde were successfully confirmed by immunofluorescence assays and transmission electron microscopy. A column was prepared and used to analyze different compounds. The results showed that the method was selective, specific, and reproducible. Overall, the high density of cells immobilized on microcarriers achieved through cell culture and covalent immobilization represents a promising strategy for affinity screening. This approach highlights the potential of the affinity screening method to identify active compounds from an herbal matrix against designed targets and its prospects for use in drug discovery from herbs.

SmartCADD: AI-QM Empowered Drug Discovery Platform with Explainability

Artificial intelligence (AI) has emerged as a pivotal force in enhancing productivity across various sectors, with its impact being profoundly felt within the pharmaceutical and biotechnology domains. Despite AI's rapid adoption, its integration into scientific research faces resistance due to myriad challenges: the opaqueness of AI models, the intricate nature of their implementation, and the issue of data scarcity. In response to these impediments, we introduce SmartCADD, an innovative, open-source virtual screening platform that combines deep learning, computer-aided drug design (CADD), and quantum mechanics methodologies within a user-friendly Python framework. SmartCADD is engineered to streamline the construction of comprehensive virtual screening workflows that incorporate a variety of formerly independent techniques─spanning ADMET property predictions, de novo 2D and 3D pharmacophore modeling, molecular docking, to the integration of explainable AI mechanisms. This manuscript highlights the foundational principles, key functionalities, and the unique integrative approach of SmartCADD. Furthermore, we demonstrate its efficacy through a case study focused on the identification of promising lead compounds for HIV inhibition. By democratizing access to advanced AI and quantum mechanics tools, SmartCADD stands as a catalyst for progress in pharmaceutical research and development, heralding a new era of innovation and efficiency.

Hydrophobic deep eutectic solvent (HDES) as oil phase in lipid-based drug formulations

There is increasing pharmaceutical interest in deep eutectic solvents not only as a green alternative to organic solvents in drug manufacturing, but also as liquid formulation for drug delivery. The present work introduces a hydrophobic deep eutectic solvent (HDES) to the field of lipid-based formulations (LBF). Phase behavior of a mixture with 2:1 M ratio of decanoic- to dodecanoic acid was studied experimentally and described by thermodynamic modelling. Venetoclax was selected as a hydrophobic model drug and studied by atomistic molecular dynamics simulations of the mixtures. As a result, valuable molecular insights were gained into the interaction networks between the different components. Moreover, experimentally the HDES showed greatly enhanced drug solubilization compared to conventional glyceride-based vehicles, but aqueous dispersion behavior was limited. Hence surfactants were studied for their ability to improve aqueous dispersion and addition of Tween 80 resulted in lowest droplet sizes and high in vitro drug release. In conclusion, the combination of HDES with surfactant(s) provides a novel LBF with high pharmaceutical potential. However, the components must be finely balanced to keep the integrity of the solubilizing HDES, while enabling sufficient dispersion and drug release.

Structural and Physicochemical Features of Oral PROTACs

Achieving oral bioavailability with Proteolysis Targeting Chimeras (PROTACs) is a key challenge. Here, we report the in vivo pharmacokinetic properties in mouse, rat, and dog of four clinical oral PROTACs and compare with an internally derived data set. We use NMR to determine 3D molecular conformations and structural preorganization free in solution, and we introduce the new experimental descriptors, solvent-exposed H-bond donors (eHBD), and acceptors (eHBA). We derive an upper limit of eHBD ≤ 2 for oral PROTACs in apolar environments and show a greater tolerance for other properties (eHBA, polarity, lipophilicity, and molecular weight) than for Rule-of-5 compliant oral drugs. Within a set of structurally related PROTACs, we show that examples with eHBD > 2 have much lower oral bioavailability than those that have eHBD ≤ 2. We summarize our findings as an experimental "Rule-of-oral-PROTACs" in order to assist medicinal chemists to achieve oral bioavailability in this challenging space.

The Pan-Canadian Chemical Library: A Mechanism to Open Academic Chemistry to High-Throughput Virtual Screening

Computationally screening chemical libraries to discover molecules with desired properties is a common technique used in early-stage drug discovery. Recent progress in the field now enables the efficient exploration of billions of molecules within days or hours, but this exploration remains confined within the boundaries of the accessible chemistry space. While the number of commercially available compounds grows rapidly, it remains a limited subset of all druglike small molecules that could be synthesized. Here, we present a workflow where chemical reactions typically developed in academia and unconventional in drug discovery are exploited to dramatically expand the chemistry space accessible to virtual screening. We use this process to generate a first version of the Pan-Canadian Chemical Library, a collection of nearly 150 billion diverse compounds that does not overlap with other ultra-large libraries such as Enamine REAL or SAVI and could be a resource of choice for protein targets where other libraries have failed to deliver bioactive molecules.

Identifying Substructures That Facilitate Compounds to Penetrate the Blood-Brain Barrier via Passive Transport Using Machine Learning Explainer Models

The local interpretable model-agnostic explanation (LIME) method was used to interpret two machine learning models of compounds penetrating the blood-brain barrier. The classification models, Random Forest, ExtraTrees, and Deep Residual Network, were trained and validated using the blood-brain barrier penetration dataset, which shows the penetrability of compounds in the blood-brain barrier. LIME was able to create explanations for such penetrability, highlighting the most important substructures of molecules that affect drug penetration in the barrier. The simple and intuitive outputs prove the applicability of this explainable model to interpreting the permeability of compounds across the blood-brain barrier in terms of molecular features. LIME explanations were filtered with a weight equal to or greater than 0.1 to obtain only the most relevant explanations. The results showed several structures that are important for blood-brain barrier penetration. In general, it was found that some compounds with nitrogenous substructures are more likely to permeate the blood-brain barrier. The application of these structural explanations may help the pharmaceutical industry and potential drug synthesis research groups to synthesize active molecules more rationally.

Plant antibacterials: The challenges and opportunities

Nature possesses an inexhaustible reservoir of agents that could serve as alternatives to combat the growing threat of antimicrobial resistance (AMR). While some of the most effective drugs for treating bacterial infections originate from natural sources, they have predominantly been derived from fungal and bacterial species. However, a substantial body of literature is available on the promising antibacterial properties of plant-derived compounds. In this comprehensive review, we address the major challenges associated with the discovery and development of plant-derived antimicrobial compounds, which have acted as obstacles preventing their clinical use. These challenges encompass limited sourcing, the risk of agent rediscovery, suboptimal drug metabolism, and pharmacokinetics (DMPK) properties, as well as a lack of knowledge regarding molecular targets and mechanisms of action, among other pertinent issues. Our review underscores the significance of these challenges and their implications in the quest for the discovery and development of effective plant-derived antimicrobial agents. Through a critical examination of the current state of research, we give valuable insights that will advance our understanding of these classes of compounds, offering potential solutions to the global crisis of AMR. © 2017 Elsevier Inc. All rights reserved.

Fragment-based drug discovery for disorders of the central nervous system: designing better drugs piece by piece

Fragment-based drug discovery (FBDD) has emerged as a powerful strategy to confront the challenges faced by conventional drug development approaches, particularly in the context of central nervous system (CNS) disorders. FBDD involves the screening of libraries that comprise thousands of small molecular fragments, each no greater than 300 Da in size. Unlike the generally larger molecules from high-throughput screening that limit customisation, fragments offer a more strategic starting point. These fragments are inherently compact, providing a strong foundation with good binding affinity for the development of drug candidates. The minimal elaboration required to transition the hit into a drug-like molecule is not only accelerated, but also it allows for precise modifications to enhance both their activity and pharmacokinetic properties. This shift towards a fragment-centric approach has seen commercial success and holds considerable promise in the continued streamlining of the drug discovery and development process. In this review, we highlight how FBDD can be integrated into the CNS drug discovery process to enhance the exploration of a target. Furthermore, we provide recent examples where FBDD has been an integral component in CNS drug discovery programs, enabling the improvement of pharmacokinetic properties that have previously proven challenging. The FBDD optimisation process provides a systematic approach to explore this vast chemical space, facilitating the discovery and design of compounds piece by piece that are capable of modulating crucial CNS targets.

Compound Absorption in Polymer Devices Impairs the Translatability of Preclinical Safety Assessments

Organotypic and microphysiological systems (MPS) that can emulate the molecular phenotype and function of human tissues, such as liver, are increasingly used in preclinical drug development. However, despite their improved predictivity, drug development success rates have remained low with most compounds failing in clinical phases despite promising preclinical data. Here, it is tested whether absorption of small molecules to polymers commonly used for MPS fabrication can impact preclinical pharmacological and toxicological assessments and contribute to the high clinical failure rates. To this end, identical devices are fabricated from eight different MPS polymers and absorption of prototypic compounds with different physicochemical properties are analyzed. It is found that overall absorption is primarily driven by compound hydrophobicity and the number of rotatable bonds. However, absorption can differ by >1000-fold between polymers with polydimethyl siloxane (PDMS) being most absorptive, whereas polytetrafluoroethylene (PTFE) and thiol-ene epoxy (TEE) absorbed the least. Strikingly, organotypic primary human liver cultures successfully flagged hydrophobic hepatotoxins in lowly absorbing TEE devices at therapeutically relevant concentrations, whereas isogenic cultures in PDMS devices are resistant, resulting in false negative safety signals. Combined, these results can guide the selection of MPS materials and facilitate the development of preclinical assays with improved translatability.

Contextualising the developability risk of antibodies with lambda light chains using enhanced therapeutic antibody profiling

Antibodies with lambda light chains (λ-antibodies) are generally considered to be less developable than those with kappa light chains (κ-antibodies). Though this hypothesis has not been formally established, it has led to substantial systematic biases in drug discovery pipelines and thus contributed to kappa dominance amongst clinical-stage therapeutics. However, the identification of increasing numbers of epitopes preferentially engaged by λ-antibodies shows there is a functional cost to neglecting to consider them as potential lead candidates. Here, we update our Therapeutic Antibody Profiler (TAP) tool to use the latest data and machine learning-based structure prediction, and apply it to evaluate developability risk profiles for κ-antibodies and λ-antibodies based on their surface physicochemical properties. We find that while human λ-antibodies on average have a higher risk of developability issues than κ-antibodies, a sizeable proportion are assigned lower-risk profiles by TAP and should represent more tractable candidates for therapeutic development. Through a comparative analysis of the low- and high-risk populations, we highlight opportunities for strategic design that TAP suggests would enrich for more developable λ-antibodies. Overall, we provide context to the differing developability of κ- and λ-antibodies, enabling a rational approach to incorporate more diversity into the initial pool of immunotherapeutic candidates.

One-Pot Synthesis of Cereblon Proteolysis Targeting Chimeras via Photoinduced C(sp2)-C(sp3) Cross Coupling and Amide Formation for Proteolysis Targeting Chimera Library Synthesis

In this study, a one-pot synthesis via photoinduced C(sp2)-C(sp3) coupling followed by amide formation to access proteolysis targeting chimeras (PROTACs) was developed. The described protocol was studied on cereblon (CRBN)-based E3-ligase binders and (+)-JQ-1, a bromodomain inhibitor, to generate BET (bromodomain and extra terminal domain) targeting protein degraders. The generated PROTACs were profiled in vitro and tested for their degradation ability with several potent candidates identified. Upfront, the individual reactions of the one-pot transformation were carefully optimized for CRBN binder functionalization and multiple heterobifunctional linker moieties were designed and synthesized. Separate scopes detailing the C(sp2)-C(sp3) coupling and one-pot PROTAC synthesis are described in this report as well as a library miniaturization study showing the high-throughput compatibility. Overall, the developed protocol provides rapid access to PROTACs in a single process, thereby allowing efficient generation of CRBN-based PROTAC libraries.

Towards holistic Compound Quality Scores: Extending ligand efficiency indices with compound pharmacokinetic characteristics

The suitability of small molecules as oral drugs is often assessed by simple physicochemical rules, the application of ligand efficiency scores or by composite scores based on physicochemical compound properties. These rules and scores are empirical and typically lack mechanistic background, such as information on pharmacokinetics (PK). We introduce new types of Compound Quality Scores (CQS, specifically called dose scores and cmax scores), which explicitly include predicted or, when available, experimental PK parameters and combine these with on-target potency. These CQS scores are surrogates for an estimated dose and corresponding cmax and allow prioritizing of compounds within test cascades as well as before synthesis. We demonstrate the complementarity and, in most cases, superior performance relative to existing efficiency metrics by project examples.

Extracting medicinal chemistry intuition via preference machine learning

The lead optimization process in drug discovery campaigns is an arduous endeavour where the input of many medicinal chemists is weighed in order to reach a desired molecular property profile. Building the expertise to successfully drive such projects collaboratively is a very time-consuming process that typically spans many years within a chemist's career. In this work we aim to replicate this process by applying artificial intelligence learning-to-rank techniques on feedback that was obtained from 35 chemists at Novartis over the course of several months. We exemplify the usefulness of the learned proxies in routine tasks such as compound prioritization, motif rationalization, and biased de novo drug design. Annotated response data is provided, and developed models and code made available through a permissive open-source license.

Expanding Bioactive Fragment Space with the Generated Database GDB-13s

Identifying innovative fragments for drug design can help medicinal chemistry address new targets and overcome the limitations of the classical molecular series. By deconstructing molecules into ring fragments (RFs, consisting of ring atoms plus ring-adjacent atoms) and acyclic fragments (AFs, consisting of only acyclic atoms), we find that public databases of molecules (i.e., ZINC and PubChem) and natural products (i.e., COCONUT) contain mostly RFs and AFs of up to 13 atoms. We also find that many RFs and AFs are enriched in bioactive vs inactive compounds from ChEMBL. We then analyze the generated database GDB-13s, which enumerates 99 million possible molecules of up to 13 atoms, for RFs and AFs resembling ChEMBL bioactive RFs and AFs. This analysis reveals a large number of novel RFs and AFs that are structurally simple, have favorable synthetic accessibility scores, and represent opportunities for synthetic chemistry to contribute to drug innovation in the context of fragment-based drug discovery.

Identification of histidine kinase inhibitors through screening of natural compounds to combat mastitis caused by Streptococcus agalactiae in dairy cattle

BACKGROUND

Mastitis poses a major threat to dairy farms globally; it results in reduced milk production, increased treatment costs, untimely compromised genetic potential, animal deaths, and economic losses. Streptococcus agalactiae is a highly virulent bacteria that cause mastitis. The administration of antibiotics for the treatment of this infection is not advised due to concerns about the emergence of antibiotic resistance and potential adverse effects on human health. Thus, there is a critical need to identify new therapeutic approaches to combat mastitis. One promising target for the development of antibacterial therapies is the transmembrane histidine kinase of bacteria, which plays a key role in signal transduction pathways, secretion systems, virulence, and antibiotic resistance.

RESULTS

In this study, we aimed to identify novel natural compounds that can inhibit transmembrane histidine kinase. To achieve this goal, we conducted a virtual screening of 224,205 natural compounds, selecting the top ten based on their lowest binding energy and favorable protein-ligand interactions. Furthermore, molecular docking of eight selected antibiotics and five histidine kinase inhibitors with transmembrane histidine kinase was performed to evaluate the binding energy with respect to top-screened natural compounds. We also analyzed the ADMET properties of these compounds to assess their drug-likeness. The top two compounds (ZINC000085569031 and ZINC000257435291) and top-screened antibiotics (Tetracycline) that demonstrated a strong binding affinity were subjected to molecular dynamics simulations (100 ns), free energy landscape, and binding free energy calculations using the MM-PBSA method.

CONCLUSION

Our results suggest that the selected natural compounds have the potential to serve as effective inhibitors of transmembrane histidine kinase and can be utilized for the development of novel antibacterial veterinary medicine for mastitis after further validation through clinical studies.

Small molecule protein kinase inhibitors approved by regulatory agencies outside of the United States

Owing to genetic alterations and overexpression, the dysregulation of protein kinases plays a significant role in the pathogenesis of many autoimmune and neoplastic disorders and protein kinase antagonists have become an important drug target. Although the efficacy of imatinib in the treatment of chronic myelogenous leukemia in the United States in 2001 was the main driver of protein kinase inhibitor drug discovery, this was preceded by the approval of fasudil (a ROCK antagonist) in Japan in 1995 for the treatment of cerebral vasospasm. There are 21 small molecule protein kinase inhibitors that are approved in China, Japan, Europe, and South Korea that are not approved in the United Sates and 75 FDA-approved inhibitors in the United States. Of the 21 agents, eleven target receptor protein-tyrosine kinases, eight inhibit nonreceptor protein-tyrosine kinases, and two block protein-serine/threonine kinases. All 21 drugs are orally bioavailable or topically effective. Of the non-FDA approved drugs, sixteen are prescribed for the treatment of neoplastic diseases, three are directed toward inflammatory disorders, one is used for glaucoma, and fasudil is used in the management of vasospasm. The leading targets of kinase inhibitors approved by both international regulatory agencies and by the FDA are members of the EGFR family, the VEGFR family, and the JAK family. One-third of the 21 internationally approved drugs are not compliant with Lipinski's rule of five for orally bioavailable drugs. The rule of five relies on four parameters including molecular weight, number of hydrogen bond donors and acceptors, and the Log of the partition coefficient.

Today's drug discovery and the shadow of the rule of 5

INTRODUCTION

The rule of 5 developed by Lipinski et al., a landmark and prescient piece of scholarship, focused the minds of drug hunters by systematically characterizing the physical make-up of drug molecules for the first time, noting many sub-optimal compounds identified by high-throughput screening practices. Its profound influence on thinking and practices, whilst providing benefit, perhaps etched the guidelines too strongly in the minds of some drug hunters who applied the bounds too literally without understanding the implications of the underlying statistics.

AREAS COVERED

This opinion is based on recent key developments that take thinking, measurements, and standards beyond those first set out, particularly the influences of molecular weight and the understanding, measurement, and calculation of lipophilicity.

EXPERT OPINION

Techniques and technologies for physicochemical estimations set new standards. It is timely to celebrate the significance and influence of the rule of 5, whilst taking thinking to new levels with better characterizations. The shadow of the rule of 5 may be long, but it is not dark, as new measurements, predictions and principles emerge as guiding lights in the design and prioritization of higher-quality molecules redefining the meaning of beyond the rule of 5.

Impact of Applicability Domains to Generative Artificial Intelligence

Molecular generative artificial intelligence is drawing significant attention in the drug design community, with several experimentally validated proof of concepts already published. Nevertheless, generative models are known for sometimes generating unrealistic, unstable, unsynthesizable, or uninteresting structures. This calls for methods to constrain those algorithms to generate structures in drug-like portions of the chemical space. While the concept of applicability domains for predictive models is well studied, its counterpart for generative models is not yet well-defined. In this work, we empirically examine various possibilities and propose applicability domains suited for generative models. Using both public and internal data sets, we use generative methods to generate novel structures that are predicted to be actives by a corresponding quantitative structure-activity relationships model while constraining the generative model to stay within a given applicability domain. Our work looks at several applicability domain definitions, combining various criteria, such as structural similarity to the training set, similarity of physicochemical properties, unwanted substructures, and quantitative estimate of drug-likeness. We assess the structures generated from both qualitative and quantitative points of view and find that the applicability domain definitions have a strong influence on the drug-likeness of generated molecules. An extensive analysis of our results allows us to identify applicability domain definitions that are best suited for generating drug-like molecules with generative models. We anticipate that this work will help foster the adoption of generative models in an industrial context.

Identification of multidentate tyrosyl-DNA phosphodiesterase 1 (TDP1) inhibitors that simultaneously access the DNA, protein and catalytic-binding sites by oxime diversification

Tyrosyl-DNA phosphodiesterase 1 (TDP1) is a member of the phospholipase D family that can downregulate the anticancer effects of the type I topoisomerase (TOP1) inhibitors by hydrolyzing the 3'-phosphodiester bond between DNA and the TOP1 residue Y723 in the critical stalled intermediate that is the foundation of TOP1 inhibitor mechanism of action. Thus, TDP1 antagonists are attractive as potential enhancers of TOP1 inhibitors. However, the open and extended nature of the TOP1-DNA substrate-binding region has made the development of TDP1 inhibitors extremely challenging. In this study, starting from our recently identified small molecule microarray (SMM)-derived TDP1-inhibitory imidazopyridine motif, we employed a click-based oxime protocol to extend the parent platform into the DNA and TOP1 peptide substrate-binding channels. We applied one-pot Groebke-Blackburn-Bienayme multicomponent reactions (GBBRs) to prepare the needed aminooxy-containing substrates. By reacting these precursors with approximately 250 aldehydes in microtiter format, we screened a library of nearly 500 oximes for their TDP1 inhibitory potencies using an in vitro florescence-based catalytic assay. Select hits were structurally explored as their triazole- and ether-based isosteres. We obtained crystal structures of two of the resulting inhibitors bound to the TDP1 catalytic domain. The structures reveal that the inhibitors form hydrogen bonds with the catalytic His-Lys-Asn triads ("HKN" motifs: H263, K265, N283 and H493, K495, N516), while simultaneously extending into both the substrate DNA and TOP1 peptide-binding grooves. This work provides a structural model for developing multivalent TDP1 inhibitors capable of binding in a tridentate fashion with a central component situated within the catalytic pocket and extensions that project into both the DNA and TOP1 peptide substrate-binding regions.

Rule of five violations among the FDA-approved small molecule protein kinase inhibitors

Because genetic alterations including mutations, overexpression, translocations, and dysregulation of protein kinases are involved in the pathogenesis of many illnesses, this enzyme family is the target of many drug discovery programs in the pharmaceutical industry. Overall, the US FDA has approved 74 small molecule protein kinase inhibitors, nearly all of which are orally effective. Of the 74 approved drugs, forty block receptor protein-tyrosine kinases, eighteen target nonreceptor protein-tyrosine kinases, twelve are directed against protein-serine/threonine protein kinases, and four target dual specificity protein kinases. The data indicate that 63 of these medicinals are approved for the management of neoplasms (51 against solid tumors such as breast, colon, and lung cancers, eight against nonsolid tumors such as leukemia, and four against both types of tumors). Seven of the FDA-approved kinase inhibitors form covalent bonds with their target enzymes and they are accordingly classified as TCIs (targeted covalent inhibitors). Medicinal chemists have examined the physicochemical properties of drugs that are orally effective. Lipinski's rule of five (Ro5) is a computational procedure that is used to estimate solubility, membrane permeability, and pharmacological effectiveness in the drug-discovery setting. It relies on four parameters including molecular weight, number of hydrogen bond donors and acceptors, and the Log of the partition coefficient. Other important descriptors include the lipophilic efficiency, the polar surface area, and the number of rotatable bonds and aromatic rings. We tabulated these and other properties of the FDA-approved kinase inhibitors. Of the 74 approved drugs, 30 fail to comply with the rule of five.

Antimicrobial Evaluation of New Pyrazoles, Indazoles and Pyrazolines Prepared in Continuous Flow Mode

Multi-drug resistant bacterial strains (MDR) have become an increasing challenge to our health system, resulting in multiple classical antibiotics being clinically inactive today. As the de-novo development of effective antibiotics is a very costly and time-consuming process, alternative strategies such as the screening of natural and synthetic compound libraries is a simple approach towards finding new lead compounds. We thus report on the antimicrobial evaluation of a small collection of fourteen drug-like compounds featuring indazoles, pyrazoles and pyrazolines as key heterocyclic moieties whose synthesis was achieved in continuous flow mode. It was found that several compounds possessed significant antibacterial potency against clinical and MDR strains of the Staphylococcus and Enterococcus genera, with the lead compound (9) reaching MIC values of 4 µg/mL on those species. In addition, time killing experiments performed on compound 9 on Staphylococcus aureus MDR strains highlight its activity as bacteriostatic. Additional evaluations regarding the physiochemical and pharmacokinetic properties of the most active compounds are reported and showcased, promising drug-likeness, which warrants further explorations of the newly identified antimicrobial lead compound.