Artificial Intelligence in ADME Property Prediction

Absorption, distribution, metabolism, excretion (ADME) are key properties of a small molecule that govern pharmacokinetic profiles and impact its efficacy and safety. Computational methods such as machine learning and artificial intelligence have gained significant interest in both academic and industrial settings to predict pharmacokinetic properties of small molecules. These methods are applied in drug discovery to optimize chemical libraries, prioritize hits from biological screens, and optimize ADME properties of lead molecules. In the recent years, the drug discovery community witnessed the use of a range of neural network architectures such as deep neural networks, recurrent neural networks, graph neural networks, and transformer neural networks, which marked a paradigm shift in computer-aided drug design and development. This chapter discusses recent developments with an emphasis on their application to predict ADME properties.

Consideration of vendor-related differences in hepatic metabolic stability data to optimize early ADME screening in drug discovery

Hepatic metabolic stability is a crucial determinant of oral bioavailability and plasma concentrations of a compound, and its measurement is important in early drug discovery. Preliminary metabolic stability estimations are commonly performed in liver microsomal fractions. At the National Center for Advancing Translational Sciences, a single-point assay in rat liver microsomes (RLM) is employed for initial stability assessment (Tier I) and a multi-point detailed stability assay is employed as a Tier II assay for promising compounds. Although the in vitro and in vivo metabolic stability of compounds typically exhibit good correlation, conflicting results may arise in certain cases. While investigating one such instance, we serendipitously found vendor-related RLM differences in metabolic stability and metabolite formation, which had implications for in vitro and in vivo correlations. In this study, we highlight the importance of considering vendor differences in hepatic metabolic stability data and discuss strategies to avoid these pitfalls.

Development and validation of PAMPA-BBB QSAR model to predict brain penetration potential of novel drug candidates

Efficiently circumventing the blood-brain barrier (BBB) poses a major hurdle in the development of drugs that target the central nervous system. Although there are several methods to determine BBB permeability of small molecules, the Parallel Artificial Membrane Permeability Assay (PAMPA) is one of the most common assays in drug discovery due to its robust and high-throughput nature. Drug discovery is a long and costly venture, thus, any advances to streamline this process are beneficial. In this study, ∼2,000 compounds from over 60 NCATS projects were screened in the PAMPA-BBB assay to develop a quantitative structure-activity relationship model to predict BBB permeability of small molecules. After analyzing both state-of-the-art and latest machine learning methods, we found that random forest based on RDKit descriptors as additional features provided the best training balanced accuracy (0.70 ± 0.015) and a message-passing variant of graph convolutional neural network that uses RDKit descriptors provided the highest balanced accuracy (0.72) on a prospective validation set. Finally, we correlated in vitro PAMPA-BBB data with in vivo brain permeation data in rodents to observe a categorical correlation of 77%, suggesting that models developed using data from PAMPA-BBB can forecast in vivo brain permeability. Given that majority of prior research has relied on in vitro or in vivo data for assessing BBB permeability, our model, developed using the largest PAMPA-BBB dataset to date, offers an orthogonal means to estimate BBB permeability of small molecules. We deposited a subset of our data into PubChem bioassay database (AID: 1845228) and deployed the best performing model on the NCATS Open Data ADME portal (https://opendata.ncats.nih.gov/adme/). These initiatives were undertaken with the aim of providing valuable resources for the drug discovery community.

The application of acute oral toxicity computational models in dangerous goods classification

Acute oral toxicity (AOT) data inform the acute toxicity potential of a compound and guides occupational safety and transportation practices. AOT data enable the categorization of a chemical into the appropriate AOT Globally Harmonized System (GHS) category based on the severity of the hazard. AOT data are also utilized to identify compounds that are Dangerous Goods (DGs) and subsequent transportation guidance for shipping of these hazardous materials. Proper identification of DGs is challenging for novel compounds that lack data. It is not feasible to err on the side of caution for all compounds lacking AOT data and to designate them as DGs, as shipping a compound as a DG has cost, resource, and time implications. With the wealth of available historical AOT data, AOT testing approaches are evolving, and in silico AOT models are emerging as tools that can be utilized with confidence to assess the acute toxicity potential of de novo molecules. Such approaches align with the 3R principles, offering a reduction or even replacement of traditional in vivo testing methods and can also be leveraged for product stewardship purposes. Utilizing proprietary historical in vivo AOT data for 210 pharmaceutical compounds (PCs), we evaluated the performance of two established in silico AOT programs: the Leadscope AOT Model Suite and the Collaborative Acute Toxicity Modeling Suite. These models accurately identified 94% and 97% compounds that were not DGs (GHS categories 4, 5, and not classified (NC)) suggesting that the models are fit-for-purpose in identifying PCs with low acute oral toxicity potential (LD50 >300 mg/kg). Utilization of these models to identify compounds that are not DGs can enable them to be de-prioritized for in vivo testing. This manuscript provides a detailed evaluation and assessment of the two models and recommends the most suitable applications of such models.

A Workflow of Integrated Resources to Catalyze Network Pharmacology Driven COVID-19 Research

In the event of an outbreak due to an emerging pathogen, time is of the essence to contain or to mitigate the spread of the disease. Drug repositioning is one of the strategies that has the potential to deliver therapeutics relatively quickly. The SARS-CoV-2 pandemic has shown that integrating critical data resources to drive drug-repositioning studies, involving host-host, host-pathogen, and drug-target interactions, remains a time-consuming effort that translates to a delay in the development and delivery of a life-saving therapy. Here, we describe a workflow we designed for a semiautomated integration of rapidly emerging data sets that can be generally adopted in a broad network pharmacology research setting. The workflow was used to construct a COVID-19 focused multimodal network that integrates 487 host-pathogen, 63 278 host-host protein, and 1221 drug-target interactions. The resultant Neo4j graph database named "Neo4COVID19" is made publicly accessible via a web interface and via API calls based on the Bolt protocol. Details for accessing the database are provided on a landing page (https://neo4covid19.ncats.io/). We believe that our Neo4COVID19 database will be a valuable asset to the research community and will catalyze the discovery of therapeutics to fight COVID-19.

NCATS Inxight Drugs: a comprehensive and curated portal for translational research

The United States has a complex regulatory scheme for marketing drugs. Understanding drug regulatory status is a daunting task that requires integrating data from many sources from the United States Food and Drug Administration (FDA), US government publications, and other processes related to drug development. At NCATS, we created Inxight Drugs (https://drugs.ncats.io), a web resource that attempts to address this challenge in a systematic manner. NCATS Inxight Drugs incorporates and unifies a wealth of data, including those supplied by the FDA and from independent public sources. The database offers a substantial amount of manually curated literature data unavailable from other sources. Currently, the database contains 125 036 product ingredients, including 2566 US approved drugs, 6242 marketed drugs, and 9684 investigational drugs. All substances are rigorously defined according to the ISO 11238 standard to comply with existing regulatory standards for unique drug substance identification. A special emphasis was placed on capturing manually curated and referenced data on treatment modalities and semantic relationships between substances. A supplementary resource 'Novel FDA Drug Approvals' features regulatory details of newly approved FDA drugs. The database is regularly updated using NCATS Stitcher data integration tool that automates data aggregation and supports full data access through a RESTful API.

How to Illuminate the Druggable Genome Using Pharos

Pharos is an integrated web-based informatics platform for the analysis of data aggregated by the Illuminating the Druggable Genome (IDG) Knowledge Management Center, an NIH Common Fund initiative. The current version of Pharos (as of October 2019) spans 20,244 proteins in the human proteome, 19,880 disease and phenotype associations, and 226,829 ChEMBL compounds. This resource not only collates and analyzes data from over 60 high-quality resources to generate these types, but also uses text indexing to find less apparent connections between targets, and has recently begun to collaborate with institutions that generate data and resources. Proteins are ranked according to a knowledge-based classification system, which can help researchers to identify less studied "dark" targets that could be potentially further illuminated. This is an important process for both drug discovery and target validation, as more knowledge can accelerate target identification, and previously understudied proteins can serve as novel targets in drug discovery. Two basic protocols illustrate the levels of detail available for targets and several methods of finding targets of interest. An Alternate Protocol illustrates the difference in available knowledge between less and more studied targets. © 2020 by John Wiley & Sons, Inc. Basic Protocol 1: Search for a target and view details Alternate Protocol: Search for dark target and view details Basic Protocol 2: Filter a target list to get refined results.

Large-Scale Modeling of Multispecies Acute Toxicity End Points Using Consensus of Multitask Deep Learning Methods

Computational methods to predict molecular properties regarding safety and toxicology represent alternative approaches to expedite drug development, screen environmental chemicals, and thus significantly reduce associated time and costs. There is a strong need and interest in the development of computational methods that yield reliable predictions of toxicity, and many approaches, including the recently introduced deep neural networks, have been leveraged towards this goal. Herein, we report on the collection, curation, and integration of data from the public data sets that were the source of the ChemIDplus database for systemic acute toxicity. These efforts generated the largest publicly available such data set comprising > 80,000 compounds measured against a total of 59 acute systemic toxicity end points. This data was used for developing multiple single- and multitask models utilizing random forest, deep neural networks, convolutional, and graph convolutional neural network approaches. For the first time, we also reported the consensus models based on different multitask approaches. To the best of our knowledge, prediction models for 36 of the 59 end points have never been published before. Furthermore, our results demonstrated a significantly better performance of the consensus model obtained from three multitask learning approaches that particularly predicted the 29 smaller tasks (less than 300 compounds) better than other models developed in the study. The curated data set and the developed models have been made publicly available at https://github.com/ncats/ld50-multitask, https://predictor.ncats.io/, and https://cactus.nci.nih.gov/download/acute-toxicity-db (data set only) to support regulatory and research applications.

TCRD and Pharos 2021: mining the human proteome for disease biology

In 2014, the National Institutes of Health (NIH) initiated the Illuminating the Druggable Genome (IDG) program to identify and improve our understanding of poorly characterized proteins that can potentially be modulated using small molecules or biologics. Two resources produced from these efforts are: The Target Central Resource Database (TCRD) (http://juniper.health.unm.edu/tcrd/) and Pharos (https://pharos.nih.gov/), a web interface to browse the TCRD. The ultimate goal of these resources is to highlight and facilitate research into currently understudied proteins, by aggregating a multitude of data sources, and ranking targets based on the amount of data available, and presenting data in machine learning ready format. Since the 2017 release, both TCRD and Pharos have produced two major releases, which have incorporated or expanded an additional 25 data sources. Recently incorporated data types include human and viral-human protein-protein interactions, protein-disease and protein-phenotype associations, and drug-induced gene signatures, among others. These aggregated data have enabled us to generate new visualizations and content sections in Pharos, in order to empower users to find new areas of study in the druggable genome.

Critical Assessment of Artificial Intelligence Methods for Prediction of hERG Channel Inhibition in the "Big Data" Era

The rise of novel artificial intelligence (AI) methods necessitates their benchmarking against classical machine learning for a typical drug-discovery project. Inhibition of the potassium ion channel, whose alpha subunit is encoded by the human ether-à-go-go-related gene (hERG), leads to a prolonged QT interval of the cardiac action potential and is a significant safety pharmacology target for the development of new medicines. Several computational approaches have been employed to develop prediction models for the assessment of hERG liabilities of small molecules including recent work using deep learning methods. Here, we perform a comprehensive comparison of hERG effect prediction models based on classical approaches (random forests and gradient boosting) and modern AI methods [deep neural networks (DNNs) and recurrent neural networks (RNNs)]. The training set (∼9000 compounds) was compiled by integrating the hERG bioactivity data from the ChEMBL database with experimental data generated from an in-house, high-throughput thallium flux assay. We utilized different molecular descriptors including the latent descriptors, which are real-value continuous vectors derived from chemical autoencoders trained on a large chemical space (>1.5 million compounds). The models were prospectively validated on ∼840 in-house compounds screened in the same thallium flux assay. The best results were obtained with the XGBoost method and RDKit descriptors. The comparison of models based only on latent descriptors revealed that the DNNs performed significantly better than the classical methods. The RNNs that operate on SMILES provided the highest model sensitivity. The best models were merged into a consensus model that offered superior performance compared to reference models from academic and commercial domains. Furthermore, we shed light on the potential of AI methods to exploit the big data in chemistry and generate novel chemical representations useful in predictive modeling and tailoring a new chemical space.

Retrospective assessment of rat liver microsomal stability at NCATS: data and QSAR models

Hepatic metabolic stability is a key pharmacokinetic parameter in drug discovery. Metabolic stability is usually assessed in microsomal fractions and only the best compounds progress in the drug discovery process. A high-throughput single time point substrate depletion assay in rat liver microsomes (RLM) is employed at the National Center for Advancing Translational Sciences. Between 2012 and 2020, RLM stability data was generated for ~ 24,000 compounds from more than 250 projects that cover a wide range of pharmacological targets and cellular pathways. Although a crucial endpoint, little or no data exists in the public domain. In this study, computational models were developed for predicting RLM stability using different machine learning methods. In addition, a retrospective time-split validation was performed, and local models were built for projects that performed poorly with global models. Further analysis revealed inherent medicinal chemistry knowledge potentially useful to chemists in the pursuit of synthesizing metabolically stable compounds. In addition, we deposited experimental data for ~ 2500 compounds in the PubChem bioassay database (AID: 1508591). The global prediction models are made publicly accessible ( https://opendata.ncats.nih.gov/adme ). This is to the best of our knowledge, the first publicly available RLM prediction model built using high-quality data generated at a single laboratory.

A Workflow of Integrated Resources to Catalyze Network Pharmacology Driven COVID-19 Research

MOTIVATION

In the event of an outbreak due to an emerging pathogen, time is of the essence to contain or to mitigate the spread of the disease. Drug repositioning is one of the strategies that has the potential to deliver therapeutics relatively quickly. The SARS-CoV-2 pandemic has shown that integrating critical data resources to drive drug-repositioning studies, involving host-host, hostpathogen and drug-target interactions, remains a time-consuming effort that translates to a delay in the development and delivery of a life-saving therapy.

RESULTS

Here, we describe a workflow we designed for a semi-automated integration of rapidly emerging datasets that can be generally adopted in a broad network pharmacology research setting. The workflow was used to construct a COVID-19 focused multimodal network that integrates 487 host-pathogen, 74,805 host-host protein and 1,265 drug-target interactions. The resultant Neo4j graph database named "Neo4COVID19" is accessible via a web interface and via API calls based on the Bolt protocol. We believe that our Neo4COVID19 database will be a valuable asset to the research community and will catalyze the discovery of therapeutics to fight COVID-19.

AVAILABILITY

https://neo4covid19.ncats.io.

Predicting liver cytosol stability of small molecules

Over the last few decades, chemists have become skilled at designing compounds that avoid cytochrome P (CYP) 450 mediated metabolism. Typical screening assays are performed in liver microsomal fractions and it is possible to overlook the contribution of cytosolic enzymes until much later in the drug discovery process. Few data exist on cytosolic enzyme-mediated metabolism and no reliable tools are available to chemists to help design away from such liabilities. In this study, we screened 1450 compounds for liver cytosol-mediated metabolic stability and extracted transformation rules that might help medicinal chemists in optimizing compounds with these liabilities. In vitro half-life data were collected by performing in-house experiments in mouse (CD-1 male) and human (mixed gender) cytosol fractions. Matched molecular pairs analysis was performed in conjunction with qualitative-structure activity relationship modeling to identify chemical structure transformations affecting cytosolic stability. The transformation rules were prospectively validated on the test set. In addition, selected rules were validated on a diverse chemical library and the resulting pairs were experimentally tested to confirm whether the identified transformations could be generalized. The validation results, comprising nearly 250 library compounds and corresponding half-life data, are made publicly available. The datasets were also used to generate in silico classification models, based on different molecular descriptors and machine learning methods, to predict cytosol-mediated liabilities. To the best of our knowledge, this is the first systematic in silico effort to address cytosolic enzyme-mediated liabilities.

SuperDRUG2: a one stop resource for approved/marketed drugs

Regular monitoring of drug regulatory agency web sites and similar resources for information on new drug approvals and changes to legal status of marketed drugs is impractical. It requires navigation through several resources to find complete information about a drug as none of the publicly accessible drug databases provide all features essential to complement in silico drug discovery. Here, we propose SuperDRUG2 (http://cheminfo.charite.de/superdrug2) as a comprehensive knowledge-base of approved and marketed drugs. We provide the largest collection of drugs (containing 4587 active pharmaceutical ingredients) which include small molecules, biological products and other drugs. The database is intended to serve as a one-stop resource providing data on: chemical structures, regulatory details, indications, drug targets, side-effects, physicochemical properties, pharmacokinetics and drug–drug interactions. We provide a 3D-superposition feature that facilitates estimation of the fit of a drug in the active site of a target with a known ligand bound to it. Apart from multiple other search options, we introduced pharmacokinetics simulation as a unique feature that allows users to visualise the ‘plasma concentration versus time’ profile for a given dose of drug with few other adjustable parameters to simulate the kinetics in a healthy individual and poor or extensive metabolisers.

Pain-Prescription Differences - An Analysis of 500,000 Discharge Summaries

BACKGROUND

Pain-relief prescriptions have led to an alarming increase in drug-related abuse.

OBJECTIVE

In this study, we estimate the pain reliever prescription rates at a major German academic hospital center and compare with the nationwide trends from Germany and prescription reports from the USA.

METHODS

We analysed >500,000 discharge summaries from Charité, encompassing the years 2006 to 2015, and extracted the medications and diagnoses from each discharge summary. Prescription reports from the USA and Germany were collected and compared with the trends at Charité to identify the frequently prescribed pain relievers and their world-wide utilization trends. The average costs of pain therapy were also calculated and compared between the three regions.

RESULTS

Metamizole (dipyrone), a non-opioid analgesic, was the most commonly prescribed pain reliever at Charité (59%) and in Germany (23%) while oxycodone (29%), a semi-synthetic opioid, was most commonly ordered in the USA. Surprisingly, metamizole was prescribed to nearly 20% of all patients at Charité, a drug that has been banned for safety reasons (agranulocytosis) in most developed countries including Canada, United Kingdom, and USA. A large number of prospective cases with high risk for agranulocytosis and other side effects were found. The average cost of pain therapy greatly varied between the USA (125.3 EUR) and Charité (17.2 EUR).

CONCLUSION

The choice of pain relievers varies regionally and is often in disagreement with approved indications and regulatory guidelines. A pronounced East-West gradient was observed with metamizole use and the opposite with prescription opioids.

The Catch-22 of Predicting hERG Blockade Using Publicly Accessible Bioactivity Data

Drug-induced inhibition of the human ether-à-go-go-related gene (hERG)-encoded potassium ion channels can lead to fatal cardiotoxicity. Several marketed drugs and promising drug candidates were recalled because of this concern. Diverse modeling methods ranging from molecular similarity assessment to quantitative structure-activity relationship analysis employing machine learning techniques have been applied to data sets of varying size and composition (number of blockers and nonblockers). In this study, we highlight the challenges involved in the development of a robust classifier for predicting the hERG end point using bioactivity data extracted from the public domain. To this end, three different modeling methods, nearest neighbors, random forests, and support vector machines, were employed to develop predictive models using different molecular descriptors, activity thresholds, and training set compositions. Our models demonstrated superior performance in external validations in comparison with those reported in the previous studies from which the data sets were extracted. The choice of descriptors had little influence on the model performance, with minor exceptions. The criteria used to filter bioactivity data, the activity threshold settings used to separate blockers from nonblockers, and the structural diversity of blockers in training data set were found to be the crucial indicators of model performance. Training sets based on a binary threshold of 1 μM/10 μM to separate blockers (IC₅₀/ K_i ≤ 1 μM) from nonblockers (IC₅₀/ K_i > 10 μM) provided superior performance in comparison with those defined using a single threshold (1 μM or 10 μM). A major limitation in using the public domain hERG activity data is the abundance of blockers in comparison with nonblockers at usual activity thresholds, since not many studies report the latter.

Drugs as habitable planets in the space of dark chemical matter

A recent study demonstrated antifungal activity of dark chemical matter (DCM) compounds that were otherwise inactive in more than 100 HTS assays. These compounds were proposed to possess unique activity and 'clean' safety profiles. Here, we present an outlook of the promiscuity and safety of these compounds by retrospectively comparing their chemical and biological spaces with those of drugs. Significant amounts of marketed drugs (16%), withdrawn drugs (16.5%) and natural compounds (3.5%) share structural identity with DCM. Compound promiscuity assessment indicates that dark matter compounds could potentially interact with multiple biological targets. Further, thousands of DCM compounds showed presence of frequent-hitting pan-assay interference compound (PAINS) substructures. In light of these observations, filtering these compounds from screening libraries can be an irrevocable loss.

Computational methods for prediction of in vitro effects of new chemical structures

BACKGROUND

With a constant increase in the number of new chemicals synthesized every year, it becomes important to employ the most reliable and fast in silico screening methods to predict their safety and activity profiles. In recent years, in silico prediction methods received great attention in an attempt to reduce animal experiments for the evaluation of various toxicological endpoints, complementing the theme of replace, reduce and refine. Various computational approaches have been proposed for the prediction of compound toxicity ranging from quantitative structure activity relationship modeling to molecular similarity-based methods and machine learning. Within the "Toxicology in the 21st Century" screening initiative, a crowd-sourcing platform was established for the development and validation of computational models to predict the interference of chemical compounds with nuclear receptor and stress response pathways based on a training set containing more than 10,000 compounds tested in high-throughput screening assays.

RESULTS

Here, we present the results of various molecular similarity-based and machine-learning based methods over an independent evaluation set containing 647 compounds as provided by the Tox21 Data Challenge 2014. It was observed that the Random Forest approach based on MACCS molecular fingerprints and a subset of 13 molecular descriptors selected based on statistical and literature analysis performed best in terms of the area under the receiver operating characteristic curve values. Further, we compared the individual and combined performance of different methods. In retrospect, we also discuss the reasons behind the superior performance of an ensemble approach, combining a similarity search method with the Random Forest algorithm, compared to individual methods while explaining the intrinsic limitations of the latter.

CONCLUSIONS

Our results suggest that, although prediction methods were optimized individually for each modelled target, an ensemble of similarity and machine-learning approaches provides promising performance indicating its broad applicability in toxicity prediction.

WITHDRAWN--a resource for withdrawn and discontinued drugs

Post-marketing drug withdrawals can be associated with various events, ranging from safety issues such as reported deaths or severe side-effects, to a multitude of non-safety problems including lack of efficacy, manufacturing, regulatory or business issues. During the last century, the majority of drugs voluntarily withdrawn from the market or prohibited by regulatory agencies was reported to be related to adverse drug reactions. Understanding the underlying mechanisms of toxicity is of utmost importance for current and future drug discovery. Here, we present WITHDRAWN, a resource for withdrawn and discontinued drugs publicly accessible at http://cheminfo.charite.de/withdrawn. Today, the database comprises 578 withdrawn or discontinued drugs, their structures, important physico-chemical properties, protein targets and relevant signaling pathways. A special focus of the database lies on the drugs withdrawn due to adverse reactions and toxic effects. For approximately one half of the drugs in the database, safety issues were identified as the main reason for withdrawal. Withdrawal reasons were extracted from the literature and manually classified into toxicity types representing adverse effects on different organs. A special feature of the database is the presence of multiple search options which will allow systematic analyses of withdrawn drugs and their mechanisms of toxicity.