A preclinical secondary pharmacology resource illuminates target-adverse drug reaction associations of marketed drugs

Machine Learning for Toxicity Prediction Using Chemical Structures: Pillars for Success in the Real World

Machine learning (ML) is increasingly valuable for predicting molecular properties and toxicity in drug discovery. However, toxicity-related end points have always been challenging to evaluate experimentally with respect to in vivo translation due to the required resources for human and animal studies; this has impacted data availability in the field. ML can augment or even potentially replace traditional experimental processes depending on the project phase and specific goals of the prediction. For instance, models can be used to select promising compounds for on-target effects or to deselect those with undesirable characteristics (e.g., off-target or ineffective due to unfavorable pharmacokinetics). However, reliance on ML is not without risks, due to biases stemming from nonrepresentative training data, incompatible choice of algorithm to represent the underlying data, or poor model building and validation approaches. This might lead to inaccurate predictions, misinterpretation of the confidence in ML predictions, and ultimately suboptimal decision-making. Hence, understanding the predictive validity of ML models is of utmost importance to enable faster drug development timelines while improving the quality of decisions. This perspective emphasizes the need to enhance the understanding and application of machine learning models in drug discovery, focusing on well-defined data sets for toxicity prediction based on small molecule structures. We focus on five crucial pillars for success with ML-driven molecular property and toxicity prediction: (1) data set selection, (2) structural representations, (3) model algorithm, (4) model validation, and (5) translation of predictions to decision-making. Understanding these key pillars will foster collaboration and coordination between ML researchers and toxicologists, which will help to advance drug discovery and development.

The Missing Enzymes: A Call to Update Pharmacological Profiling Practices for Better Drug Safety Assessment

Pharmacological profiling is critical for the development of safe drugs. With increasing awareness of its significance and attempts to share best practices, here we aimed to understand how pharmacological profiling is implemented and reported in the primary literature by analyzing the representation of nonkinase enzymes in selectivity screens. This aspect has been overlooked in previous publications, despite enzymes constituting a significant portion of the pharmacological targets for currently marketed drugs. Our analysis shows that while industry recommendations for improved pharmacological profiling have been widely adopted, enzymes remain largely underrepresented: about a quarter of studies did not include enzymes, and on average, enzymes comprise only 11% of all targets in pharmacological screens. We discuss possible reasons for this shortcoming and provide examples of critical enzymes missing from current screens. We conclude with the notion that selectivity screens should be expanded to include more enzymes to improve drug development and safety.

Computational drug repurposing: approaches, evaluation of in silico resources and case studies

Repurposing of existing drugs for new indications has attracted substantial attention owing to its potential to accelerate drug development and reduce costs. Hundreds of computational resources such as databases and predictive platforms have been developed that can be applied for drug repurposing, making it challenging to select the right resource for a specific drug repurposing project. With the aim of helping to address this challenge, here we overview computational approaches to drug repurposing based on a comprehensive survey of available in silico resources using a purpose-built drug repurposing ontology that classifies the resources into hierarchical categories and provides application-specific information. We also present an expert evaluation of selected resources and three drug repurposing case studies implemented within the Horizon Europe REMEDi4ALL project to demonstrate the practical use of the resources. This comprehensive Review with expert evaluations and case studies provides guidelines and recommendations on the best use of various in silico resources for drug repurposing and establishes a basis for a sustainable and extendable drug repurposing web catalogue.

Transformer Decoder Learns from a Pretrained Protein Language Model to Generate Ligands with High Affinity

The drug discovery process can be significantly accelerated by using deep learning methods to suggest molecules with druglike features and, more importantly, that are good candidates to bind specific proteins of interest. We present a novel deep learning generative model, Prot2Drug, that learns to generate ligands binding specific targets leveraging (i) the information carried by a pretrained protein language model and (ii) the ability of transformers to capitalize the knowledge gathered from thousands of protein-ligand interactions. The embedding unveils the receipt to follow for designing molecules binding a given protein, and Prot2Drug translates such instructions by using the syntax of the molecular language generating novel compounds which are predicted to have favorable physicochemical properties and high affinity toward specific targets. Moreover, Prot2Drug reproduced numerous known interactions between compounds and proteins used for generating them and suggested novel protein targets for known compounds, indicating potential drug repurposing strategies. Remarkably, Prot2Drug facilitates the design of promising ligands even for protein targets with limited or no information about their ligands or 3D structure.

Disruption of serotonin homeostasis in intestinal organoids provides insights into drug-induced gastrointestinal toxicity

Drug-induced gastrointestinal toxicity is a frequent clinical adverse event that needs to be carefully monitored and managed to ensure patient compliance. While preclinical assessment of drug-induced gastrointestinal toxicity mostly relies on animal experimentation, intestinal organoids have gained increasing attention to identify gastrointestinal toxicants in vitro. Nonetheless, current in vitro protocols primarily assess structural alterations induced by drugs, whereas gastrointestinal adverse events can often stem from functional disturbances. Disruption of serotonin signaling in the gastrointestinal tract is associated with impaired motility, as well as nausea and vomiting. We aimed to investigate alterations of serotonin homeostasis in organoids derived from the canine small intestine as a driver of drug-induced gastrointestinal toxicity. Treatment of the organoids with a compound (NVS-1) inducing acute gastrointestinal toxicity in dogs as well as with three tyrosine kinase inhibitors with known preclinical and clinical gastrointestinal adverse effects (afatinib, crizotinib and vandetanib) led to increased supernatant serotonin levels. Mechanistic assays showed that, while NVS-1 and afatinib stimulate serotonin release, crizotinib and vandetanib inhibit serotonin re-uptake via direct inhibition of the serotonin re-uptake transporter. Using a data mining approach, we further suggest that inhibition of serotonin re-uptake could contribute to gastrointestinal toxicity observed with multiple marketed drugs. In conclusion, we present the implementation of a novel in vitro gastrointestinal toxicity endpoint that could complement current methods and serve as a mechanistic and predictive/screening tool for drug-induced gastrointestinal toxicity.

Predicting clinical outcomes from off-target receptor interactions using Secondary Intelligence™

Adverse effects due to off-target activity can be predicted by careful comparison of the relationship between expected plasma concentration and off-target activity of the test compound with that of reference drugs targeting that receptor for their therapeutic efficacy. The ratio between plasma concentration (unbound) and the Ki at the receptor is a surrogate measure reflecting receptor occupancy. Where data are available for reference drugs, we have curated and evaluated this at 100 receptors, 72 of which can involve both negative and positive modulations by drugs: a total of 172 'receptor modulations'. This provides a quantitative framework upon which to achieve consistent risk assessment of off-target interactions across receptors, across compounds and between assessors. It therefore represents a significant departure from an opinion-based to an evidence-based approach to secondary pharmacology. Demonstration of proof-of-principle was achieved for one of the receptor interactions (α1A-adrenoceptor antagonism leading to postural hypotension in clinical use) due to the availability of high-quality off-target Ki data for >30 drugs at this receptor.

In vitro pharmacologic profiling aids systemic toxicity assessment of chemicals

An adapted in vitro pharmacology profiling panel (APPP) was developed that included targets used in the pharmaceutical industry alongside additional targets whose cellular functions have been linked to systemic toxicities. This panel of 83 target assays was used to profile the activities of 129 cosmetic relevant chemicals with diverse chemical structures, physiochemical properties and cosmetic use types. Internal data consistency was proved robust, as evidenced by the reproducibility between single concentration and concentration-response data and showed good concordance with data reported in the ToxCast and drug excipient datasets. We discuss how the data can be analyzed and multiple potential contexts of use illustrated by case studies, alongside other new approach methodologies, to support cosmetic chemical risk assessments that do not require data from new animal studies.

Polypharmacological profiling across protein target families and cellular pathways using the multiplexed cell-based assay platform safetyProfiler reveals efficacy, potency and side effects of drugs

Selectivity profiling is key for assessing the pharmacological properties of multi-target drugs. We have developed a cell-based and barcoded assay encompassing ten druggable targets, including G protein-coupled receptors (GPCRs), receptor tyrosine kinases (RTKs), nuclear receptors, a protease as well as their key downstream pathways and profiled 17 drugs in living cells for efficacy, potency, and side effects. Notably, this multiplex assay, termed safetyProfiler assay, enabled the simultaneous assessment of multiple target and pathway activities, shedding light on the polypharmacological profile of compounds. For example, the neuroleptics clozapine, paliperidone, and risperidone potently inhibited primary targets DRD2 and HTR2A as well as cAMP and calcium pathways. However, while paliperidone and risperidone also potently inhibited the secondary target ADRA1A and mitogen-activated protein kinase (MAPK) downstream pathways, clozapine only exhibited mild antagonistic effects on ADRA1A and lacked MAPK inhibition downstream of DRD2 and HTR2A. Furthermore, we present data on the selectivity for bazedoxifene, an estrogen receptor antagonist currently undergoing clinical phase 2 trials for breast cancer, on MAPK signaling. Additionally, precise potency data for LY2452473, an androgen receptor antagonist, that completed a phase 2 clinical trial for prostate cancer, are presented. The non-selective kinase inhibitor staurosporine was observed to potently inactivate the two RTKs EGFR and ERBB4 as well as MAPK signaling, while eliciting stress-related cAMP responses. Our findings underscore the value of comprehensive profiling in elucidating the pharmacological properties of established and novel therapeutics, thereby facilitating the development of novel multi-target drugs with enhanced efficacy and selectivity.

On the relationship between hERG inhibition and the magnitude of QTc prolongation: An in vitro to clinical translational analysis

Assessing the magnitude of QTc prolongation is crucial in drug development due to its association with Torsades de Pointes. Inhibition of the hERG channel, pivotal in cardiac repolarization, is a key factor in evaluating this risk. In this study, the relationship between hERG inhibition and QTc prolongation magnitude was investigated, with the aim to derive simple guidance on the required hERG margin to avoid a large (>20 ms) QTc prolongation.

METHODS

Data from literature and FDA sources were searched for compounds with hERG IC50 values alongside clinical QTc data with paired plasma concentrations, or compounds demonstrating a clinical concentration-QTc relationship. Relationships between hERG inhibition, hERG IC50 margin to unbound plasma Cmax, and QTc prolongation magnitude were calculated.

RESULTS

Analysis of 148 clinical QTc observations from 98 compounds revealed that compounds associated with QTc prolongation >10 ms typically exhibited hERG margins of ≤33-fold, while those exceeding 20 ms were generally associated with margins of ≤24-fold. QTc increases above 10 ms were not observed at hERG margins >100-fold. Based on 53 clinical concentration-QTc datasets, modest hERG inhibition levels of ∼4-6 % correlated with a 10 ms QTc prolongation, while ∼10-13 % inhibition corresponded to a 20 ms prolongation.

CONCLUSIONS

This study enhances understanding of the relationship between hERG inhibition and QTc prolongation magnitude, by conducting analysis across a wide range of 98 compounds. This information can be used to determine the optimal hERG margin, particularly for drug discovery projects with limited scope to completely design-out hERG activity.

Polypharmacology prediction: the long road toward comprehensively anticipating small-molecule selectivity to de-risk drug discovery

INTRODUCTION

Small molecules often bind to multiple targets, a behavior termed polypharmacology. Anticipating polypharmacology is essential for drug discovery since unknown off-targets can modulate safety and efficacy - profoundly affecting drug discovery success. Unfortunately, experimental methods to assess selectivity present significant limitations and drugs still fail in the clinic due to unanticipated off-targets. Computational methods are a cost-effective, complementary approach to predict polypharmacology.

AREAS COVERED

This review aims to provide a comprehensive overview of the state of polypharmacology prediction and discuss its strengths and limitations, covering both classical cheminformatics methods and bioinformatic approaches. The authors review available data sources, paying close attention to their different coverage. The authors then discuss major algorithms grouped by the types of data that they exploit using selected examples.

EXPERT OPINION

Polypharmacology prediction has made impressive progress over the last decades and contributed to identify many off-targets. However, data incompleteness currently limits most approaches to comprehensively predict selectivity. Moreover, our limited agreement on model assessment challenges the identification of the best algorithms - which at present show modest performance in prospective real-world applications. Despite these limitations, the exponential increase of multidisciplinary Big Data and AI hold much potential to better polypharmacology prediction and de-risk drug discovery.

Safety Implications of Modulating Nuclear Receptors: A Comprehensive Analysis from Non-Clinical and Clinical Perspectives

The unintended modulation of nuclear receptor (NR) activity by drugs can lead to toxicities amongst the endocrine, gastrointestinal, hepatic cardiovascular, and central nervous systems. While secondary pharmacology screening assays include NRs, safety risks due to unintended interactions of small molecule drugs with NRs remain poorly understood. To identify potential nonclinical and clinical safety effects resulting from functional interactions with 44 of the 48 human-expressed NRs, we conducted a systematic narrative review of the scientific literature, tissue expression data, and used curated databases (OFF-X™) (Off-X, Clarivate) to organize reported toxicities linked to the functional modulation of NRs in a tabular and machine-readable format. The top five NRs associated with the highest number of safety alerts from peer-reviewed journals, regulatory agency communications, congresses/conferences, clinical trial registries, and company communications were the Glucocorticoid Receptor (GR, 18,328), Androgen Receptor (AR, 18,219), Estrogen Receptor (ER, 12,028), Retinoic acid receptors (RAR, 10,450), and Pregnane X receptor (PXR, 8044). Toxicities associated with NR modulation include hepatotoxicity, cardiotoxicity, endocrine disruption, carcinogenicity, metabolic disorders, and neurotoxicity. These toxicities often arise from the dysregulation of receptors like Peroxisome proliferator-activated receptors (PPARα, PPARγ), the ER, PXR, AR, and GR. This dysregulation leads to various health issues, including liver enlargement, hepatocellular carcinoma, heart-related problems, hormonal imbalances, tumor growth, metabolic syndromes, and brain function impairment. Gene expression analysis using heatmaps for human and rat tissues complemented the functional modulation of NRs associated with the reported toxicities. Interestingly, certain NRs showed ubiquitous expression in tissues not previously linked to toxicities, suggesting the potential utilization of organ-specific NR interactions for therapeutic purposes.

In silico off-target profiling for enhanced drug safety assessment

Ensuring drug safety in the early stages of drug development is crucial to avoid costly failures in subsequent phases. However, the economic burden associated with detecting drug off-targets and potential side effects through in vitro safety screening and animal testing is substantial. Drug off-target interactions, along with the adverse drug reactions they induce, are significant factors affecting drug safety. To assess the liability of candidate drugs, we developed an artificial intelligence model for the precise prediction of compound off-target interactions, leveraging multi-task graph neural networks. The outcomes of off-target predictions can serve as representations for compounds, enabling the differentiation of drugs under various ATC codes and the classification of compound toxicity. Furthermore, the predicted off-target profiles are employed in adverse drug reaction (ADR) enrichment analysis, facilitating the inference of potential ADRs for a drug. Using the withdrawn drug Pergolide as an example, we elucidate the mechanisms underlying ADRs at the target level, contributing to the exploration of the potential clinical relevance of newly predicted off-target interactions. Overall, our work facilitates the early assessment of compound safety/toxicity based on off-target identification, deduces potential ADRs of drugs, and ultimately promotes the secure development of drugs.

AI for targeted polypharmacology: The next frontier in drug discovery

In drug discovery, targeted polypharmacology, i.e., targeting multiple molecular targets with a single drug, is redefining therapeutic design to address complex diseases. Pre-selected pharmacological profiles, as exemplified in kinase drugs, promise enhanced efficacy and reduced toxicity. Historically, many of such drugs were discovered serendipitously, limiting predictability and efficacy, but currently artificial intelligence (AI) offers a transformative solution. Machine learning and deep learning techniques enable modeling protein structures, generating novel compounds, and decoding their polypharmacological effects, opening an avenue for more systematic and predictive multi-target drug design. This review explores the use of AI in identifying synergistic co-targets and delineating them from anti-targets that lead to adverse effects, and then discusses advances in AI-enabled docking, generative chemistry, and proteochemometric modeling of proteome-wide compound interactions, in the context of polypharmacology. We also provide insights into challenges ahead.