Ensemble Models Based on QuBiLS-MAS Features and Shallow Learning for the Prediction of Drug-Induced Liver Toxicity: Improving Deep Learning and Traditional Approaches

BoostDILI: Extreme Gradient Boost-Powered Drug-Induced Liver Injury Prediction and Structural Alerts Generation

Over the past 60 years, drug-induced liver injury (DILI) has played a key role in the withdrawal of marketed drugs due to safety concerns. Early prediction of DILI is crucial for developing safer pharmaceuticals, yet current in vitro and in vivo testing methods are complex and cumbersome. In this study, we developed an extreme gradient boosting (XGB)-powered machine learning (ML) model for DILI prediction. Comparing various DILI prediction models is challenging because they rely on different public data sets. We comprehensively evaluated the proposed BoostDILI model to address two crucial questions: 1. Can insights derived from public data sets help in DILI prediction for Food and Drug Administration (FDA) approved drugs? 2. Can we generate structural alerts to improve the model's explainability? To address the first question, we developed a DILI prediction model using four publicly available data sets. This effort led to the creation of the BoostDILI model, which achieved a 5-fold CV accuracy of 0.70. A sequential feature selection method was employed to identify relevant descriptors. This model integrates feature-level representations derived from RDKit (12 features) and Mordred (23 features) features. Bayesian statistics was applied to identify high-performance substructures iteratively, and a structural alerts model was developed to address the second question. The developed model was further validated with two FDA-approved drug data sets, DILIst and DILIRank. The BoostDILI model offers a trustable solution for evaluating the DILI risk in preclinical research. The structural alerts help in identifying the substructures that may be responsible for DILI. The data set and the source code are available at https://github.com/Naga270588/BoostDILI.

Role of artificial intelligence in revolutionizing drug discovery

The application of artificial intelligence (AI) in medicine, particularly through machine learning (ML), marked a significant progression in drug discovery. AI acts as a powerful catalyst in narrowing the gap between disease understanding and the identification of potential therapeutic agents. This review provides an inclusive summary of the latest advancements in AI and its application in drug discovery. We examine the various stages of the drug discovery process, starting from disease identification and encompassing diagnosis, target identification, screening, and lead discovery. AI's capability to analyze extensive datasets and discern patterns is essential in these stages, enhancing predictions and efficiencies in disease identification, drug discovery, and clinical trial management. The role of AI in expediting drug development is emphasized, highlighting its potential to analyze vast data volumes, thus reducing the time and costs associated with new drug market introduction. The importance of data quality, algorithm training, and ethical considerations, especially in patient data handling during clinical trials, is addressed. By considering these factors, AI promises to transform drug development, offering significant benefits to patients and society.

Psychedelic Drugs in Mental Disorders: Current Clinical Scope and Deep Learning-Based Advanced Perspectives

Mental disorders are a representative type of brain disorder, including anxiety, major depressive depression (MDD), and autism spectrum disorder (ASD), that are caused by multiple etiologies, including genetic heterogeneity, epigenetic dysregulation, and aberrant morphological and biochemical conditions. Psychedelic drugs such as psilocybin and lysergic acid diethylamide (LSD) have been renewed as fascinating treatment options and have gradually demonstrated potential therapeutic effects in mental disorders. However, the multifaceted conditions of psychiatric disorders resulting from individuality, complex genetic interplay, and intricate neural circuits impact the systemic pharmacology of psychedelics, which disturbs the integration of mechanisms that may result in dissimilar medicinal efficiency. The precise prescription of psychedelic drugs remains unclear, and advanced approaches are needed to optimize drug development. Here, recent studies demonstrating the diverse pharmacological effects of psychedelics in mental disorders are reviewed, and emerging perspectives on structural function, the microbiota-gut-brain axis, and the transcriptome are discussed. Moreover, the applicability of deep learning is highlighted for the development of drugs on the basis of big data. These approaches may provide insight into pharmacological mechanisms and interindividual factors to enhance drug discovery and development for advanced precision medicine.

Discovering New Tyrosinase Inhibitors by Using In Silico Modelling, Molecular Docking, and Molecular Dynamics

Background/Objectives: This study was used in silico modelling to search for potential tyrosinase protein inhibitors from a database of different core structures for IC50 prediction. Methods: Four machine learning algorithms and topographical descriptors were tested for model construction. Results: A model based on multiple linear regression was the most robust, with only six descriptors, and validated by the Tropsha test with statistical parameters R2 = 0.8687, Q2LOO = 0.8030, and Q2ext = 0.9151. From the screening of FDA-approved drugs and natural products, the pIC50 values for 15,424 structures were calculated. The applicability domain analysis covered 100% of the external dataset and 71.22% and 73.26% of the two screening datasets. Fifteen candidates with pIC50 above 7.6 were identified, with five structures proposed as potential tyrosinase enzyme inhibitors, which underwent ADME analysis. Conclusions: The molecular docking analysis was performed for the dataset used in the training-test process and for the fifteen structures from the screening dataset with potential pharmaceutical tyrosinase inhibition, followed by molecular dynamics studies for the top five candidates with the highest predicted pIC50 values. The new use of these five candidates in tyrosinase inhibition is highlighted based on their promising application in melanoma treatment.

Using computer modeling to find new LRRK2 inhibitors for parkinson's disease

Parkinson's disease (PD) is a complex neurodegenerative disorder that affects multiple neurotransmitters, and its exact cause is still unknown. Developing new drugs for PD is a lengthy and expensive process, making it difficult to find new treatments. This study aims to create a detailed dataset to build strong predictive models with various machine learning algorithms. An ensemble modeling approach was employed to screen the DrugBank database, aiming to repurpose approved medications as potential treatments for Parkinson's disease (PD). The dataset was constructed using pIC50 values of various compounds targeting the inhibition of leucine-rich repeat kinase 2 (LRRK2). The best ensemble model showed exceptional predictive performance, with five-fold cross-validation and external validation metrics exceeding 0.8 (Q2cv = 0.864 and Q2ext = 0.873). The DrugBank screening resulted in three promising drugs-triamterene, phenazopyridine, and CRA_1801-with predicted pIC50 values greater than 7, warranting further investigation as novel PD treatments. Molecular docking and molecular dynamics simulations were performed to provide a comprehensive understanding of the interactions between LRRK2 and the inhibitors in the data set and best molecules of the screening. Free energy of binding calculation along with hydrogen bond occupancy analysis and RMSD of the ligand in the pocket show CRA_1801 as the best candidate to be repurposed as LRRK2 inhibitor.

Molecular designing of potential environmentally friendly PFAS based on deep learning and generative models

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are widely used across a spectrum of industrial and consumer goods. Nonetheless, their persistent nature and tendency to accumulate in biological systems pose substantial environmental and health threats. Consequently, striking a balance between maximizing product efficiency and minimizing environmental and health risks by tailoring the molecular structure of PFAS has become a pivotal challenge in the fields of environmental chemistry and sustainable development. To address this issue, a computational workflow was proposed for designing an environmentally friendly PFAS by incorporating deep learning (DL) and molecular generative models. The hybrid DL architecture MolHGT+ based on heterogeneous graph neural network with transformer-like attention was applied to predict the surface tension, bioaccumulation, and hepatotoxicity of the molecules. Through virtual screening of the PFAS master database using MolHGT+, the findings indicate that incorporating the siloxane group and betaine fragment can effectively decrease both the bioaccumulation and hepatotoxicity of PFAS while preserving low surface tension. In addition, molecular generative models were employed to create a structurally diverse pool of novel PFASs with the aforementioned hit molecules serving as the initial template structures. Overall, our study presents a promising AI-driven method for advancing the development of environmentally friendly PFAS.

Computational approaches for lead compound discovery in dipeptidyl peptidase-4 inhibition using machine learning and molecular dynamics techniques

The prediction of possible lead compounds from already-known drugs that may present DPP-4 inhibition activity imply a advantage in the drug development in terms of time and cost to find alternative medicines for the treatment of Type 2 Diabetes Mellitus (T2DM). The inhibition of dipeptidyl peptidase-4 (DPP-4) has been one of the most explored strategies to develop potential drugs against this condition. A diverse dataset of molecules with known experimental inhibitory activity against DPP-4 was constructed and used to develop predictive models using different machine-learning algorithms. Model M36 is the most promising one based on the internal and external performance showing values of Q2CV = 0.813, and Q2EXT = 0.803. The applicability domain evaluation and Tropsha's analysis were conducted to validate M36, indicating its robustness and accuracy in predicting pIC50 values for organic molecules within the established domain. The physicochemical properties of the ligands, including electronegativity, polarizability, and van der Waals volume were relevant to predict the inhibition process. The model was then employed in the virtual screening of potential DPP4 inhibitors, finding 448 compounds from the DrugBank and 9 from DiaNat with potential inhibitory activity. Molecular docking and molecular dynamics simulations were used to get insight into the ligand-protein interaction. From the screening and the favorable molecular dynamic results, several compounds including Skimmin (pIC50 = 3.54, Binding energy = -8.86 kcal/mol), bergenin (pIC50 = 2.69, Binding energy = -13.90 kcal/mol), and DB07272 (pIC50 = 3.97, Binding energy = -25.28 kcal/mol) seem to be promising hits to be tested and optimized in the treatment of T2DM. This results imply a important reduction in cost and time on the application of this drugs because all the information about the its metabolism is already available.

Comprehensive hepatotoxicity prediction: ensemble model integrating machine learning and deep learning

Background

Chemicals may lead to acute liver injuries, posing a serious threat to human health. Achieving the precise safety profile of a compound is challenging due to the complex and expensive testing procedures. In silico approaches will aid in identifying the potential risk of drug candidates in the initial stage of drug development and thus mitigating the developmental cost.

Methods

In current studies, QSAR models were developed for hepatotoxicity predictions using the ensemble strategy to integrate machine learning (ML) and deep learning (DL) algorithms using various molecular features. A large dataset of 2588 chemicals and drugs was randomly divided into training (80%) and test (20%) sets, followed by the training of individual base models using diverse machine learning or deep learning based on three different kinds of descriptors and fingerprints. Feature selection approaches were employed to proceed with model optimizations based on the model performance. Hybrid ensemble approaches were further utilized to determine the method with the best performance.

Results

The voting ensemble classifier emerged as the optimal model, achieving an excellent prediction accuracy of 80.26%, AUC of 82.84%, and recall of over 93% followed by bagging and stacking ensemble classifiers method. The model was further verified by an external test set, internal 10-fold cross-validation, and rigorous benchmark training, exhibiting much better reliability than the published models.

Conclusion

The proposed ensemble model offers a dependable assessment with a good performance for the prediction regarding the risk of chemicals and drugs to induce liver damage.

Improved Detection of Drug-Induced Liver Injury by Integrating Predicted In Vivo and In Vitro Data

Drug-induced liver injury (DILI) has been a significant challenge in drug discovery, often leading to clinical trial failures and necessitating drug withdrawals. Over the last decade, the existing suite of in vitro proxy-DILI assays has generally improved at identifying compounds with hepatotoxicity. However, there is considerable interest in enhancing the in silico prediction of DILI because it allows for evaluating large sets of compounds more quickly and cost-effectively, particularly in the early stages of projects. In this study, we aim to study ML models for DILI prediction that first predict nine proxy-DILI labels and then use them as features in addition to chemical structural features to predict DILI. The features include in vitro (e.g., mitochondrial toxicity, bile salt export pump inhibition) data, in vivo (e.g., preclinical rat hepatotoxicity studies) data, pharmacokinetic parameters of maximum concentration, structural fingerprints, and physicochemical parameters. We trained DILI-prediction models on 888 compounds from the DILI data set (composed of DILIst and DILIrank) and tested them on a held-out external test set of 223 compounds from the DILI data set. The best model, DILIPredictor, attained an AUC-PR of 0.79. This model enabled the detection of the top 25 toxic compounds (2.68 LR+, positive likelihood ratio) compared to models using only structural features (1.65 LR+ score). Using feature interpretation from DILIPredictor, we identified the chemical substructures causing DILI and differentiated cases of DILI caused by compounds in animals but not in humans. For example, DILIPredictor correctly recognized 2-butoxyethanol as nontoxic in humans despite its hepatotoxicity in mice models. Overall, the DILIPredictor model improves the detection of compounds causing DILI with an improved differentiation between animal and human sensitivity and the potential for mechanism evaluation. DILIPredictor required only chemical structures as input for prediction and is publicly available at https://broad.io/DILIPredictor for use via web interface and with all code available for download.

Prediction of Drug-Induced Liver Injury: From Molecular Physicochemical Properties and Scaffold Architectures to Machine Learning Approaches

The process of developing new drugs is widely acknowledged as being time-intensive and requiring substantial financial investment. Despite ongoing efforts to reduce time and expenses in drug development, ensuring medication safety remains an urgent problem. One of the major problems involved in drug development is hepatotoxicity, specifically known as drug-induced liver injury (DILI). The popularity of new drugs often poses a significant barrier during development and frequently leads to their recall after launch. In silico methods have many advantages compared with traditional in vivo and in vitro assays. To establish a more precise and reliable prediction model, it is necessary to utilize an extensive and high-quality database consisting of information on drug molecule properties and structural patterns. In addition, we should also carefully select appropriate molecular descriptors that can be used to accurately depict compound characteristics. The aim of this study was to conduct a comprehensive investigation into the prediction of DILI. First, we conducted a comparative analysis of the physicochemical properties of extensively well-prepared DILI-positive and DILI-negative compounds. Then, we used classic substructure dissection methods to identify structural pattern differences between these two different types of chemical molecules. These findings indicate that it is not feasible to establish property or substructure-based rules for distinguishing between DILI-positive and DILI-negative compounds. Finally, we developed quantitative classification models for predicting DILI using the naïve Bayes classifier (NBC) and recursive partitioning (RP) machine learning techniques. The optimal DILI prediction model was obtained using NBC, which combines 21 physicochemical properties, the VolSurf descriptors and the LCFP_10 fingerprint set. This model achieved a global accuracy (GA) of 0.855 and an area under the curve (AUC) of 0.704 for the training set, while the corresponding values were 0.619 and 0.674 for the test set, respectively. Moreover, indicative substructural fragments favorable or unfavorable for DILI were identified from the best naïve Bayesian classification model. These findings may help prioritize lead compounds in the early stage of drug development pipelines.

Analysis of pharmacokinetic profile and ecotoxicological character of cefepime and its photodegradation products

An important problem is the impact of photodegradation on product toxicity in biological tests, which may be complex and context-dependent. Previous studies have described the pharmacology of cefepime, but the toxicological effects of its photodegradation products remain largely unknown. Therefore, photodegradation studies were undertaken in conditions similar to those occurring in biological systems insilico, in vitro, in vivo and ecotoxicological experiments. The structures of four cefepime photodegradation products were determined by UPLC-MS/MS method. The calculated in silico ADMET profile indicates that carcinogenic potential is expected for compounds CP-1, cefepime, CP-2 and CP-3. The Cell Line Cytomotovity Predictor 2.0 tool was used to predict the cytotoxic effects of cefepime and related compounds in non-transformed and cancer cell lines. The results indicate that possible actions include: non-small cell lung cancer, breast adenocarcinoma, prostate cancer and papillary renal cell carcinoma. OPERA models were used to predict absorption, distribution, metabolism and excretion (ADME) endpoints, and potential bioactivity of CP-2, cefepime and CP-4. The results obtained in silico show that after 96h of exposure, cefepime, CP-1, CP-2, and CP-3 are moderately toxic in the zebrafish model, while CP-4 is highly toxic. On the contrary, cefepime is more toxic to T. platyurus (highly toxic) compared to the zebrafish model, similar to products CP-4, CP-3 and CP-2. In vitro cytotoxicity studies were performed by MTT assay and in vivo acute embryo toxicity studies using Danio rerio embryos and larvae. In vitro showed an increase in the cytotoxicity of products with the longest exposure period i.e. for 8 h. Additionally, at a concentration of 200 μg/mL, statistically significant changes in metabolic activity were observed depending on the irradiation time. In vivo studies conducted with Zebrafish showed that both cefepime and its photodegradation products have only low toxicity. Assessment of potential ecotoxicity included Microbiotests on invertebrates (Thamnotoxkit F and Daphtoxkit F), and luminescence inhibition tests (LumiMara). The observed toxicity of the tested solutions towards both Thamnocephalus platyurus and Daphnia magna indicates that the parent substance (unexposed) has lower toxicity, which increases during irradiation. The acute toxicity (Lumi Mara) of nonirradiated cefepime solution is low for all tested strains (<10%), but mixtures of cefepime and its photoproducts showed growth inhibition against all tested strains (except #6, Photobacterium phoreum). Generally, it can be concluded that after UV-Vis irradiation, the mixture of cefepime phototransformation products shows a significant increase in toxicity.

Computational models for predicting liver toxicity in the deep learning era

Drug-induced liver injury (DILI) is a severe adverse reaction caused by drugs and may result in acute liver failure and even death. Many efforts have centered on mitigating risks associated with potential DILI in humans. Among these, quantitative structure-activity relationship (QSAR) was proven to be a valuable tool for early-stage hepatotoxicity screening. Its advantages include no requirement for physical substances and rapid delivery of results. Deep learning (DL) made rapid advancements recently and has been used for developing QSAR models. This review discusses the use of DL in predicting DILI, focusing on the development of QSAR models employing extensive chemical structure datasets alongside their corresponding DILI outcomes. We undertake a comprehensive evaluation of various DL methods, comparing with those of traditional machine learning (ML) approaches, and explore the strengths and limitations of DL techniques regarding their interpretability, scalability, and generalization. Overall, our review underscores the potential of DL methodologies to enhance DILI prediction and provides insights into future avenues for developing predictive models to mitigate DILI risk in humans.

GeoDILI: A Robust and Interpretable Model for Drug-Induced Liver Injury Prediction Using Graph Neural Network-Based Molecular Geometric Representation

Drug-induced liver injury (DILI) is a significant cause of drug failure and withdrawal due to liver damage. Accurate prediction of hepatotoxic compounds is crucial for safe drug development. Several DILI prediction models have been published, but they are built on different data sets, making it difficult to compare model performance. Moreover, most existing models are based on molecular fingerprints or descriptors, neglecting molecular geometric properties and lacking interpretability. To address these limitations, we developed GeoDILI, an interpretable graph neural network that uses a molecular geometric representation. First, we utilized a geometry-based pretrained molecular representation and optimized it on the DILI data set to improve predictive performance. Second, we leveraged gradient information to obtain high-precision atomic-level weights and deduce the dominant substructure. We benchmarked GeoDILI against recently published DILI prediction models, as well as popular GNN models and fingerprint-based machine learning models using the same data set, showing superior predictive performance of our proposed model. We applied the interpretable method in the DILI data set and derived seven precise and mechanistically elucidated structural alerts. Overall, GeoDILI provides a promising approach for accurate and interpretable DILI prediction with potential applications in drug discovery and safety assessment. The data and source code are available at GitHub repository (https://github.com/CSU-QJY/GeoDILI).

Collaborative SAR Modeling and Prospective In Vitro Validation of Oxidative Stress Activation in Human HepG2 Cells

Oxidative stress is the consequence of an abnormal increase of reactive oxygen species (ROS). ROS are generated mainly during the metabolism in both normal and pathological conditions as well as from exposure to xenobiotics. Xenobiotics can, on the one hand, disrupt molecular machinery involved in redox processes and, on the other hand, reduce the effectiveness of the antioxidant activity. Such dysregulation may lead to oxidative damage when combined with oxidative stress overpassing the cell capacity to detoxify ROS. In this work, a green fluorescent protein (GFP)-tagged nuclear factor erythroid 2-related factor 2 (NRF2)-regulated sulfiredoxin reporter (Srxn1-GFP) was used to measure the antioxidant response of HepG2 cells to a large series of drug and drug-like compounds (2230 compounds). These compounds were then classified as positive or negative depending on cellular response and distributed among different modeling groups to establish structure-activity relationship (SAR) models. A selection of models was used to prospectively predict oxidative stress induced by a new set of compounds subsequently experimentally tested to validate the model predictions. Altogether, this exercise exemplifies the different challenges of developing SAR models of a phenotypic cellular readout, model combination, chemical space selection, and results interpretation.

Artificial Intelligence in Drug Toxicity Prediction: Recent Advances, Challenges, and Future Perspectives

Toxicity prediction is a critical step in the drug discovery process that helps identify and prioritize compounds with the greatest potential for safe and effective use in humans, while also reducing the risk of costly late-stage failures. It is estimated that over 30% of drug candidates are discarded owing to toxicity. Recently, artificial intelligence (AI) has been used to improve drug toxicity prediction as it provides more accurate and efficient methods for identifying the potentially toxic effects of new compounds before they are tested in human clinical trials, thus saving time and money. In this review, we present an overview of recent advances in AI-based drug toxicity prediction, including the use of various machine learning algorithms and deep learning architectures, of six major toxicity properties and Tox21 assay end points. Additionally, we provide a list of public data sources and useful toxicity prediction tools for the research community and highlight the challenges that must be addressed to enhance model performance. Finally, we discuss future perspectives for AI-based drug toxicity prediction. This review can aid researchers in understanding toxicity prediction and pave the way for new methods of drug discovery.

ElectroPredictor: An Application to Predict Mayr's Electrophilicity E through Implementation of an Ensemble Model Based on Machine Learning Algorithms

Electrophilicity (E) is one of the most important parameters to understand the reactivity of an organic molecule. Although the theoretical electrophilicity index (ω) has been associated with E in a small homologous series, the use of w to predict E in a structurally heterogeneous set of compounds is not a trivial task. In this study, a robust ensemble model is created using Mayr's database of reactivity parameters. A combination of topological and quantum mechanical descriptors and different machine learning algorithms are employed for the model's development. The predictability of the model is assessed using different statistical parameters, and its validation is examined, including a training/test partition, an applicability domain, and a y-scrambling test. The global ensemble model presents a Q5-fold2 of 0.909 and a Qext2 of 0.912, demonstrating an excellent predictability performance of E values and showing that w is not a good descriptor for the prediction of E, especially for the case of neutral compounds. ElectroPredictor, a noncommercial Python application (https://github.com/mmoreno1/ElectroPredictor), is developed to predict E. QM9, a well-known large dataset containing 133885 neutral molecules, is used to perform a virtual screening (94.0% coverage). Finally, the 10 most electrophilic molecules are analyzed as possible new Mayr's electrophiles, which have not yet been experimentally tested. This study confirms the necessity to build an ensemble model using nonlinear machine learning algorithms, topographic descriptors, and separating molecules into charged and neutral compounds to predict E with precision.

Searching glycolate oxidase inhibitors based on QSAR, molecular docking, and molecular dynamic simulation approaches

Primary hyperoxaluria type 1 (PHT1) treatment is mainly focused on inhibiting the enzyme glycolate oxidase, which plays a pivotal role in the production of glyoxylate, which undergoes oxidation to produce oxalate. When the renal secretion capacity exceeds, calcium oxalate forms stones that accumulate in the kidneys. In this respect, detailed QSAR analysis, molecular docking, and dynamics simulations of a series of inhibitors containing glycolic, glyoxylic, and salicylic acid groups have been performed employing different regression machine learning techniques. Three robust models with less than 9 descriptors-based on a tenfold cross (Q² _CV) and external (Q² _EXT) validation-were found i.e., MLR1 (Q² _CV = 0.893, Q² _EXT = 0.897), RF1 (Q² _CV = 0.889, Q² _EXT = 0.907), and IBK1 (Q² _CV = 0.891, Q² _EXT = 0.907). An ensemble model was built by averaging the predicted pIC₅₀ of the three models, obtaining a Q² _EXT = 0.933. Physicochemical properties such as charge, electronegativity, hardness, softness, van der Waals volume, and polarizability were considered as attributes to build the models. To get more insight into the potential biological activity of the compouds studied herein, docking and dynamic analysis were carried out, finding the hydrophobic and polar residues show important interactions with the ligands. A screening of the DrugBank database V.5.1.7 was performed, leading to the proposal of seven commercial drugs within the applicability domain of the models, that can be suggested as possible PHT1 treatment.

Relationships between the Structure and Severe Drug-Induced Liver Injury for Low, Medium, and High Doses of Drugs

Assessment of structure-activity relationships (SARs) for predicting severe drug-induced liver injury (DILI) is essential since in vivo and in vitro preclinical methods cannot detect many druglike compounds disrupting liver functions. To date, plenty of SAR models for the prediction of DILI have been developed; however, none of them considered the route of drug administration and daily dose, which may introduce significant bias into prediction results. We have created a dataset of 617 drugs with parenteral and oral administration routes and consistent information on DILI severity. We have found a clear relationship between route, dose, and DILI severity. According to SAR, nearly 40% of moderate- and non-DILI-causing drugs would cause severe DILI if they were administered at high oral doses. We have proposed the following approach to predict severe DILI. New compounds recommended to be used at low oral doses (<∼10 mg daily), or parenterally, can be considered not causing severe DILI. DILI for compounds administered at medium oral doses (∼10-100 mg daily; 22.2% of drugs under consideration) can be considered unpredictable because reasonable SAR models were not obtained due to the small size and heterogeneity of the corresponding dataset. The DILI potential of the compounds recommended to be used at high oral doses (more than ∼100 mg daily) can be estimated using SAR modeling. The balanced accuracy of the approach calculated by a 10-fold cross-validation procedure is 0.803. The developed approach can be used to estimate severe DILI for druglike compounds proposed to use at low and high oral doses or parenterally at the early stages of drug development.

In Silico Searching for Alternative Lead Compounds to Treat Type 2 Diabetes through a QSAR and Molecular Dynamics Study

Free fatty acid receptor 1 (FFA1) stimulates insulin secretion in pancreatic β-cells. An advantage of therapies that target FFA1 is their reduced risk of hypoglycemia relative to common type 2 diabetes treatments. In this work, quantitative structure-activity relationship (QSAR) approach was used to construct models to identify possible FFA1 agonists by applying four different machine-learning algorithms. The best model (M2) meets the Tropsha's test requirements and has the statistics parameters R2 = 0.843, Q2CV = 0.785, and Q2ext = 0.855. Also, coverage of 100% of the test set based on the applicability domain analysis was obtained. Furthermore, a deep analysis based on the ADME predictions, molecular docking, and molecular dynamics simulations was performed. The lipophilicity and the residue interactions were used as relevant criteria for selecting a candidate from the screening of the DiaNat and DrugBank databases. Finally, the FDA-approved drugs bilastine, bromfenac, and fenofibric acid are suggested as potential and lead FFA1 agonists.

Machine Learning Models for Predicting Liver Toxicity

Liver toxicity is a major adverse drug reaction that accounts for drug failure in clinical trials and withdrawal from the market. Therefore, predicting potential liver toxicity at an early stage in drug discovery is crucial to reduce costs and the potential for drug failure. However, current in vivo animal toxicity testing is very expensive and time consuming. As an alternative approach, various machine learning models have been developed to predict potential liver toxicity in humans. This chapter reviews current advances in the development and application of machine learning models for prediction of potential liver toxicity in humans and discusses possible improvements to liver toxicity prediction.

In Silico Models for Hepatotoxicity

In this chapter, we review the state of the art of predicting human hepatotoxicity using in silico techniques. There has been significant progress in this area over the past 20 years but there are still some challenges ahead. Principally, these challenges are our partial understanding of a very complex biochemical system and our ability to emulate that in a predictive capacity. Here, we provide an overview of the published modeling approaches in this area to date and discuss their design, strengths and weaknesses. It is interesting to note the diversity in modeling approaches, whether they be statistical algorithms or evidenced-based approaches including structural alerts and pharmacophore models. Irrespective of modeling approach, it appears a common theme of access to appropriate, relevant, and high-quality data is a limitation to all and is likely to continue to be the focus of future research.

Joint Decision-Making Model Based on Consensus Modeling Technology for the Prediction of Drug-Induced Liver Injury

Drug-induced liver injury (DILI) is the major cause of clinical trial failure and postmarketing withdrawals of approved drugs. It is very expensive and time-consuming to evaluate hepatotoxicity using animal or cell-based experiments in the early stage of drug development. In this study, an in silico model based on the joint decision-making strategy was developed for DILI assessment using a relatively large dataset of 2608 compounds. Five consensus models were developed with PaDEL descriptors and PubChem, Substructure, Estate, and Klekota–Roth fingerprints, respectively. Submodels for each consensus model were obtained through joint optimization. The parameters and features of each submodel were optimized jointly based on the hybrid quantum particle swarm optimization (HQPSO) algorithm. The application domain (AD) based on the frequency-weighted and distance (FWD)-based method and Tanimoto similarity index showed the wide AD of the qualified consensus models. A joint decision-making model was integrated by the qualified consensus models, and the overwhelming majority principle was used to improve the performance of consensus models. The application scope narrowing caused by the overwhelming majority principle was successfully solved by joint decision-making. The proposed model successfully predicted 99.2% of the compounds in the test set, with an accuracy of 80.0%, a sensitivity of 83.9, and a specificity of 73.3%. For an external validation set containing 390 compounds collected from DILIrank, 98.2% of the compounds were successfully predicted with an accuracy of 79.9%, a sensitivity of 97.1%, and a specificity of 66.0%. Furthermore, 25 privileged substructures responsible for DILI were identified from Substructure, PubChem, and Klekota–Roth fingerprints. These privileged substructures can be regarded as structural alerts in hepatotoxicity evaluation. Compared with the main published studies, our method exhibits certain advantage in data size, transparency, and standardization of the modeling process and accuracy and credibility of prediction results. It is a promising tool for virtual screening in the early stage of drug development.

FP-ADMET: a compendium of fingerprint-based ADMET prediction models

MOTIVATION

The absorption, distribution, metabolism, excretion, and toxicity (ADMET) of drugs plays a key role in determining which among the potential candidates are to be prioritized. In silico approaches based on machine learning methods are becoming increasing popular, but are nonetheless limited by the availability of data. With a view to making both data and models available to the scientific community, we have developed FPADMET which is a repository of molecular fingerprint-based predictive models for ADMET properties. In this article, we have examined the efficacy of fingerprint-based machine learning models for a large number of ADMET-related properties. The predictive ability of a set of 20 different binary fingerprints (based on substructure keys, atom pairs, local path environments, as well as custom fingerprints such as all-shortest paths) for over 50 ADMET and ADMET-related endpoints have been evaluated as part of the study. We find that for a majority of the properties, fingerprint-based random forest models yield comparable or better performance compared with traditional 2D/3D molecular descriptors.

AVAILABILITY

The models are made available as part of open access software that can be downloaded from https://gitlab.com/vishsoft/fpadmet .

Molecular insights on ABL kinase activation using tree-based machine learning models and molecular docking

Abelson kinase (c-Abl) is a non-receptor tyrosine kinase involved in several biological processes essential for cell differentiation, migration, proliferation, and survival. This enzyme's activation might be an alternative strategy for treating diseases such as neutropenia induced by chemotherapy, prostate, and breast cancer. Recently, a series of compounds that promote the activation of c-Abl has been identified, opening a promising ground for c-Abl drug development. Structure-based drug design (SBDD) and ligand-based drug design (LBDD) methodologies have significantly impacted recent drug development initiatives. Here, we combined SBDD and LBDD approaches to characterize critical chemical properties and interactions of identified c-Abl's activators. We used molecular docking simulations combined with tree-based machine learning models-decision tree, AdaBoost, and random forest to understand the c-Abl activators' structural features required for binding to myristoyl pocket, and consequently, to promote enzyme and cellular activation. We obtained predictive and robust models with Matthews correlation coefficient values higher than 0.4 for all endpoints and identified characteristics that led to constructing a structure-activity relationship model (SAR).

In Silico Screening of the DrugBank Database to Search for Possible Drugs against SARS-CoV-2

Coronavirus desease 2019 (COVID-19) is responsible for more than 1.80 M deaths worldwide. A Quantitative Structure-Activity Relationships (QSAR) model is developed based on experimental pIC50 values reported for a structurally diverse dataset. A robust model with only five descriptors is found, with values of R2 = 0.897, Q2LOO = 0.854, and Q2ext = 0.876 and complying with all the parameters established in the validation Tropsha's test. The analysis of the applicability domain (AD) reveals coverage of about 90% for the external test set. Docking and molecular dynamic analysis are performed on the three most relevant biological targets for SARS-CoV-2: main protease, papain-like protease, and RNA-dependent RNA polymerase. A screening of the DrugBank database is executed, predicting the pIC50 value of 6664 drugs, which are IN the AD of the model (coverage = 79%). Fifty-seven possible potent anti-COVID-19 candidates with pIC50 values > 6.6 are identified, and based on a pharmacophore modelling analysis, four compounds of this set can be suggested as potent candidates to be potential inhibitors of SARS-CoV-2. Finally, the biological activity of the compounds was related to the frontier molecular orbitals shapes.

Deep Graph Learning with Property Augmentation for Predicting Drug-Induced Liver Injury

Drug-induced liver injury (DILI) is a crucial factor in determining the qualification of potential drugs. However, the DILI property is excessively difficult to obtain due to the complex testing process. Consequently, an in silico screening in the early stage of drug discovery would help to reduce the total development cost by filtering those drug candidates with a high risk to cause DILI. To serve the screening goal, we apply several computational techniques to predict the DILI property, including traditional machine learning methods and graph-based deep learning techniques. While deep learning models require large training data to tune huge model parameters, the DILI data set only contains a few hundred annotated molecules. To alleviate the data scarcity problem, we propose a property augmentation strategy to include massive training data with other property information. Extensive experiments demonstrate that our proposed method significantly outperforms all existing baselines on the DILI data set by obtaining a 81.4% accuracy using cross-validation with random splitting, 78.7% using leave-one-out cross-validation, and 76.5% using cross-validation with scaffold splitting.

Predicting the reproductive toxicity of chemicals using ensemble learning methods and molecular fingerprints

Reproductive toxicity endpoints are a significant safety concern in the assessment of the adverse effects of chemicals in drug discovery. Computational models that can accurately predict a chemical's toxic potential are increasingly pursued to replace traditional animal experiments. Thus, ensemble learning models were built to predict the reproductive toxicity of compounds. Our ensemble models were developed using support vector machine, random forest, and extreme gradient boosting methods and 9 molecular fingerprints calculated for a dataset containing 1823 chemicals. The best prediction performance was achieved by the Ensemble-Top12 model, with an accuracy (ACC) of 86.33 %, a sensitivity (SEN) of 82.02 %, a specificity (SPE) of 90.19 %, and an area under the receiver operating characteristic curve (AUC) of 0.937 in 5-fold cross-validation and ACC, SEN, SPE, and AUC values of 84.38 %, 86.90 %, 90.67 %, and 0.920, respectively, in external validation. We also defined the applicability domain (AD) of the ensemble model by calculating the Tanimoto distance of the training set. Compared with models in existing literature, our ensemble model achieves relatively high ACC, SPE and AUC values. We also identified several fingerprint features related to chemical reproductive toxicity. Considering the performance of model, we recommend using the Ensemble-Top12 model to predict reproductive toxicity in early drug development.

QSAR and molecular docking modelling of anti-leishmanial activities of organic selenium and tellurium compounds

Leishmaniasis affects mainly rural areas and the poorest people in the world. A computational study of the antileishmanial activity of organic selenium and tellurium compounds was performed. The 3D structures of the compounds were optimized at the wb97xd/lanl2dz level and used in the quantitative structure-activity relationship (QSAR) analysis. The antileishmanial activity was measured by L. donovani β carbonic anhydrase inhibition (Ki) and the half-maximal inhibitory concentration (IC₅₀) against L. infantum amastigotes. The dataset was divided into training (75%) and test sets (25%) by using a k-means clustering algorithm. For pKi prediction, model M3 with seven 3D topographic descriptors was characterized by the following statistical parameters: r ² = 0.879, Q ² _LOO = 0.822, and Q ² _ext = 0.840. For pIC₅₀ prediction, model M12 with six attributes was characterized by the following statistical parameters: r ² = 0.907, Q ² _LOO = 0.824, and Q ² _ext = 0.795. Both models met all the requirements of Tropsha´s test, which implies predictions of pIC₅₀ and pKi activities with high accuracy. Concomitantly, favourable interactions of the sulphonamide group with the Zn atom in the protein were revealed by the docking analysis.

DeepDILI: Deep Learning-Powered Drug-Induced Liver Injury Prediction Using Model-Level Representation

Drug-induced liver injury (DILI) is the most frequently reported single cause of safety-related withdrawal of marketed drugs. It is essential to identify drugs with DILI potential at the early stages of drug development. In this study, we describe a deep learning-powered DILI (DeepDILI) prediction model created by combining model-level representation generated by conventional machine learning (ML) algorithms with a deep learning framework based on Mold2 descriptors. We conducted a comprehensive evaluation of the proposed DeepDILI model performance by posing several critical questions: (1) Could the DILI potential of newly approved drugs be predicted by accumulated knowledge of early approved ones? (2) is model-level representation more informative than molecule-based representation for DILI prediction? and (3) could improved model explainability be established? For question 1, we developed the DeepDILI model using drugs approved before 1997 to predict the DILI potential of those approved thereafter. As a result, the DeepDILI model outperformed the five conventional ML algorithms and two state-of-the-art ensemble methods with a Matthews correlation coefficient (MCC) value of 0.331. For question 2, we demonstrated that the DeepDILI model's performance was significantly improved (i.e., a MCC improvement of 25.86% in test set) compared with deep neural networks based on molecule-based representation. For question 3, we found 21 chemical descriptors that were enriched, suggesting a strong association with DILI outcome. Furthermore, we found that the DeepDILI model has more discrimination power to identify the DILI potential of drugs belonging to the World Health Organization therapeutic category of 'alimentary tract and metabolism'. Moreover, the DeepDILI model based on Mold2 descriptors outperformed the ones with Mol2vec and MACCS descriptors. Finally, the DeepDILI model was applied to the recent real-world problem of predicting any DILI concern for potential COVID-19 treatments from repositioning drug candidates. Altogether, this developed DeepDILI model could serve as a promising tool for screening for DILI risk of compounds in the preclinical setting, and the DeepDILI model is publicly available through https://github.com/TingLi2016/DeepDILI.