"DompeKeys": a set of novel substructure-based descriptors for efficient chemical space mapping, development and structural interpretation of machine learning models, and indexing of large databases.

The conversion of chemical structures into computer-readable descriptors, able to capture key structural aspects, is of pivotal importance in the field of cheminformatics and computer-aided drug design. Molecular fingerprints represent a widely employed class of descriptors; however, their generation process is time-consuming for large databases and is susceptible to bias. Therefore, descriptors able to accurately detect predefined structural fragments and devoid of lengthy generation procedures would be highly desirable. To meet additional needs, such descriptors should also be interpretable by medicinal chemists, and suitable for indexing databases with trillions of compounds. To this end, we developed-as integral part of EXSCALATE, Dompé's end-to-end drug discovery platform-the DompeKeys (DK), a new substructure-based descriptor set, which encodes the chemical features that characterize compounds of pharmaceutical interest. DK represent an exhaustive collection of curated SMARTS strings, defining chemical features at different levels of complexity, from specific functional groups and structural patterns to simpler pharmacophoric points, corresponding to a network of hierarchically interconnected substructures. Because of their extended and hierarchical structure, DK can be used, with good performance, in different kinds of applications. In particular, we demonstrate how they are very well suited for effective mapping of chemical space, as well as substructure search and virtual screening. Notably, the incorporation of DK yields highly performing machine learning models for the prediction of both compounds' activity and metabolic reaction occurrence. The protocol to generate the DK is freely available at https://dompekeys.exscalate.eu and is fully integrated with the Molecular Anatomy protocol for the generation and analysis of hierarchically interconnected molecular scaffolds and frameworks, thus providing a comprehensive and flexible tool for drug design applications.

MetaSpot: A General Approach for Recognizing the Reactive Atoms Undergoing Metabolic Reactions Based on the MetaQSAR Database.

The prediction of drug metabolism is attracting great interest for the possibility of discarding molecules with unfavorable ADME/Tox profile at the early stage of the drug discovery process. In this context, artificial intelligence methods can generate highly performing predictive models if they are trained by accurate metabolic data. MetaQSAR-based datasets were collected to predict the sites of metabolism for most metabolic reactions. The models were based on a set of structural, physicochemical, and stereo-electronic descriptors and were generated by the random forest algorithm. For each considered biotransformation, two types of models were developed: the first type involved all non-reactive atoms and included atom types among the descriptors, while the second type involved only non-reactive centers having the same atom type(s) of the reactive atoms. All the models of the first type revealed very high performances; the models of the second type show on average worst performances while being almost always able to recognize the reactive centers; only conjugations with glucuronic acid are unsatisfactorily predicted by the models of the second type. Feature evaluation confirms the major role of lipophilicity, self-polarizability, and H-bonding for almost all considered reactions. The obtained results emphasize the possibility of recognizing the sites of metabolism by classification models trained on MetaQSAR database. The two types of models can be synergistically combined since the first models identify which atoms can undergo a given metabolic reactions, while the second models detect the truly reactive centers. The generated models are available as scripts for the VEGA program.

ProfhEX: AI-based platform for small molecules liability profiling.

Off-target drug interactions are a major reason for candidate failure in the drug discovery process. Anticipating potential drug's adverse effects in the early stages is necessary to minimize health risks to patients, animal testing, and economical costs. With the constantly increasing size of virtual screening libraries, AI-driven methods can be exploited as first-tier screening tools to provide liability estimation for drug candidates. In this work we present ProfhEX, an AI-driven suite of 46 OECD-compliant machine learning models that can profile small molecules on 7 relevant liability groups: cardiovascular, central nervous system, gastrointestinal, endocrine, renal, pulmonary and immune system toxicities. Experimental affinity data was collected from public and commercial data sources. The entire chemical space comprised 289'202 activity data for a total of 210'116 unique compounds, spanning over 46 targets with dataset sizes ranging from 819 to 18896. Gradient boosting and random forest algorithms were initially employed and ensembled for the selection of a champion model. Models were validated according to the OECD principles, including robust internal (cross validation, bootstrap, y-scrambling) and external validation. Champion models achieved an average Pearson correlation coefficient of 0.84 (SD of 0.05), an R2 determination coefficient of 0.68 (SD = 0.1) and a root mean squared error of 0.69 (SD of 0.08). All liability groups showed good hit-detection power with an average enrichment factor at 5% of 13.1 (SD of 4.5) and AUC of 0.92 (SD of 0.05). Benchmarking against already existing tools demonstrated the predictive power of ProfhEX models for large-scale liability profiling. This platform will be further expanded with the inclusion of new targets and through complementary modelling approaches, such as structure and pharmacophore-based models. ProfhEX is freely accessible at the following address: https://profhex.exscalate.eu/ .

Ensemble of structure and ligand-based classification models for hERG liability profiling.

Drug-induced cardiotoxicity represents one of the most critical safety concerns in the early stages of drug development. The blockade of the human ether-à-go-go-related potassium channel (hERG) is the most frequent cause of cardiotoxicity, as it is associated to long QT syndrome which can lead to fatal arrhythmias. Therefore, assessing hERG liability of new drugs candidates is crucial to avoid undesired cardiotoxic effects. In this scenario, computational approaches have emerged as useful tools for the development of predictive models able to identify potential hERG blockers. In the last years, several efforts have been addressed to generate ligand-based (LB) models due to the lack of experimental structural information about hERG channel. However, these methods rely on the structural features of the molecules used to generate the model and often fail in correctly predicting new chemical scaffolds. Recently, the 3D structure of hERG channel has been experimentally solved enabling the use of structure-based (SB) strategies which may overcome the limitations of the LB approaches. In this study, we compared the performances achieved by both LB and SB classifiers for hERG-related cardiotoxicity developed by using Random Forest algorithm and employing a training set containing 12789 hERG binders. The SB models were trained on a set of scoring functions computed by docking and rescoring calculations, while the LB classifiers were built on a set of physicochemical descriptors and fingerprints. Furthermore, models combining the LB and SB features were developed as well. All the generated models were internally validated by ten-fold cross-validation on the TS and further verified on an external test set. The former revealed that the best performance was achieved by the LB model, while the model combining the LB and the SB attributes displayed the best results when applied on the external test set highlighting the usefulness of the integration of LB and SB features in correctly predicting unseen molecules. Overall, our predictive models showed satisfactory performances providing new useful tools to filter out potential cardiotoxic drug candidates in the early phase of drug discovery.

"Molecular Anatomy": a new multi-dimensional hierarchical scaffold analysis tool.

The scaffold representation is widely employed to classify bioactive compounds on the basis of common core structures or correlate compound classes with specific biological activities. In this paper, we present a novel approach called "Molecular Anatomy" as a flexible and unbiased molecular scaffold-based metrics to cluster large set of compounds. We introduce a set of nine molecular representations at different abstraction levels, combined with fragmentation rules, to define a multi-dimensional network of hierarchically interconnected molecular frameworks. We demonstrate that the introduction of a flexible scaffold definition and multiple pruning rules is an effective method to identify relevant chemical moieties. This approach allows to cluster together active molecules belonging to different molecular classes, capturing most of the structure activity information, in particular when libraries containing a huge number of singletons are analyzed. We also propose a procedure to derive a network visualization that allows a full graphical representation of compounds dataset, permitting an efficient navigation in the scaffold's space and significantly contributing to perform high quality SAR analysis. The protocol is freely available as a web interface at https://ma.exscalate.eu .

CATMoS: Collaborative Acute Toxicity Modeling Suite.

BACKGROUND: Humans are exposed to tens of thousands of chemical substances that need to be assessed for their potential toxicity. Acute systemic toxicity testing serves as the basis for regulatory hazard classification, labeling, and risk management. However, it is cost- and time-prohibitive to evaluate all new and existing chemicals using traditional rodent acute toxicity tests. In silico models built using existing data facilitate rapid acute toxicity predictions without using animals. OBJECTIVES: The U.S. Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) Acute Toxicity Workgroup organized an international collaboration to develop in silico models for predicting acute oral toxicity based on five different end points: Lethal Dose 50 (LD50 value, U.S. Environmental Protection Agency hazard (four) categories, Globally Harmonized System for Classification and Labeling hazard (five) categories, very toxic chemicals [LD50 (LD50≤50mg/kg)], and nontoxic chemicals (LD50>2,000mg/kg). METHODS: An acute oral toxicity data inventory for 11,992 chemicals was compiled, split into training and evaluation sets, and made available to 35 participating international research groups that submitted a total of 139 predictive models. Predictions that fell within the applicability domains of the submitted models were evaluated using external validation sets. These were then combined into consensus models to leverage strengths of individual approaches. RESULTS: The resulting consensus predictions, which leverage the collective strengths of each individual model, form the Collaborative Acute Toxicity Modeling Suite (CATMoS). CATMoS demonstrated high performance in terms of accuracy and robustness when compared with in vivo results. DISCUSSION: CATMoS is being evaluated by regulatory agencies for its utility and applicability as a potential replacement for in vivo rat acute oral toxicity studies. CATMoS predictions for more than 800,000 chemicals have been made available via the National Toxicology Program's Integrated Chemical Environment tools and data sets (ice.ntp.niehs.nih.gov). The models are also implemented in a free, standalone, open-source tool, OPERA, which allows predictions of new and untested chemicals to be made. https://doi.org/10.1289/EHP8495.

Visualization and Analysis of the REACH-chemical Space with Generative Topographic Mapping.

In the framework of REACH (Registration Evaluation Authorization and restriction of Chemicals) regulation, industries have generated and reported a huge amount of (eco)toxicological data on substance produced or imported in Europe. The registration procedure initiated the creation of a large REACH database of well defined (eco)toxicological properties. Here, the data distribution in the REACH chemical space was analyzed with the help of the Generative Topographic Mapping (GTM) approach. GTM generates 2-dimensional maps on which each compound is represented as a data point. The 3rd dimension can be used in order to display a distribution of the given (eco)toxicological property, which can further be used for property assessment of new compounds projected on the map. We report the "Universal REACH map" which accommodates 11 endpoints, covering environmental fate and (eco)toxicological properties. This map demonstrates acceptable predictive performance: in cross-validation, balanced accuracy ranges from 0.60 to 0.78. The 11 endpoints profile has been computed for each REACH-registered substance. Some concerns related to acute aquatic toxicity have been identified, whereas for environmental fate and human health endpoints the amount of compounds predicted as of concern was much smaller. It has been demonstrated that superposition of several class landscapes allows to select the zones in the chemical space populated by compounds with a given (eco)toxicological profile.

Accuracy Evaluation of Three Modelling Tools for Occupational Exposure Assessment.

OBJECTIVES: The objective of this study is to evaluate the accuracy and robustness of three exposure-modelling tools [STOFFENMANAGER® v.6, European Centre for Ecotoxicology and Toxicology of Chemical Target Risk Assessment v.3.1 (ECETOC TRA v.3.1), and Advanced REACH Tool (ART v.1.5)], by comparing available measured data for exposure to organic solvents and pesticides in occupational exposure scenarios (ESs). METHODS: Model accuracy was evaluated by comparing the predicted and the measured values, expressed as an underestimation or overestimation factor (PRED/EXP), and by regression analysis. Robustness was quantitatively described by the so-called variable 'Uncertainty Factor' (UF), which was attributed to each model's input: a higher UF score indicates greater model uncertainty and poorer robustness. RESULTS: ART was the most accurate model, with median PRED/EXP factors of 1.3 and 0.15 for organic solvent and pesticide ESs, respectively, and a significant correlation (P < 0.05) among estimated and measured data. As expected, Tier 1 model ECETOC TRA demonstrated the worst performance in terms of accuracy, with median PRED/EXP factors of 2.0 for organic solvent ESs and 3545 for pesticide ESs. Simultaneously, STOFFENMANAGER® showed a median UF equal to 2.0, resulting in the most robust model. DISCUSSION: ECETOC TRA was not considered acceptable in terms of accuracy, confirming that this model is not appropriate for the evaluation of the selected ESs for pesticides. Conversely, STOFFENMANAGER® was the best choice, and ART tended to underestimate the exposure to pesticides. For organic solvent ESs, there were no cases of strong underestimation, and all models presented overall acceptable results; for the selected ESs, ART showed the best accuracy. Stoffenmanager was the most robust model overall, indicating that even with a mistake in ES interpretation, predicted values would remain acceptable. CONCLUSION: ART may lead to more accurate results when well-documented ESs are available. In other situations, Stoffenmanager appears to be a safer alternative because of its greater robustness, particularly when entry data uncertainty is difficult to assess. ECETOC TRA cannot be directly compared to higher tiered models because of its simplistic nature: the use of this tool should be limited only to exceptional cases in which a strong conservative and worst-case evaluation is necessary.