DIGEP-Pred 2.0: A web application for predicting drug-induced cell signaling and gene expression changes.

The analysis of drug-induced gene expression profiles (DIGEP) is widely used to estimate the potential therapeutic and adverse drug effects as well as the molecular mechanisms of drug action. However, the corresponding experimental data is absent for many existing drugs and drug-like compounds. To solve this problem, we created the DIGEP-Pred 2.0 web application, which allows predicting DIGEP and potential drug targets by structural formula of drug-like compounds. It is based on the combined use of structure-activity relationships (SARs) and network analysis. SAR models were created using PASS (Prediction of Activity Spectra for Substances) technology for data from the Comparative Toxicogenomics Database (CTD), the Connectivity Map (CMap) for the prediction of DIGEP, and PubChem and ChEMBL for the prediction of molecular mechanisms of action (MoA). Using only the structural formula of a compound, the user can obtain information on potential gene expression changes in several cell lines and drug targets, which are potential master regulators responsible for the observed DIGEP. The mean accuracy of prediction calculated by leave-one-out cross validation was 86.5 % for 13377 genes and 94.8 % for 2932 proteins (CTD data), and it was 97.9 % for 2170 MoAs. SAR models (mean accuracy-87.5 %) were also created for CMap data given on MCF7, PC3, and HL60 cell lines with different threshold values for the logarithm of fold changes: 0.5, 0.7, 1, 1.5, and 2. Additionally, the data on pathways (KEGG, Reactome), biological processes of Gene Ontology, and diseases (DisGeNet) enriched by the predicted genes, together with the estimation of target-master regulators based on OmniPath data, is also provided. DIGEP-Pred 2.0 web application is freely available at https://www.way2drug.com/digep-pred.

SAnDReS 2.0: Development of machine-learning models to explore the scoring function space.

Classical scoring functions may exhibit low accuracy in determining ligand binding affinity for proteins. The availability of both protein-ligand structures and affinity data make it possible to develop machine-learning models focused on specific protein systems with superior predictive performance. Here, we report a new methodology named SAnDReS that combines AutoDock Vina 1.2 with 54 regression methods available in Scikit-Learn to calculate binding affinity based on protein-ligand structures. This approach allows exploration of the scoring function space. SAnDReS generates machine-learning models based on crystal, docked, and AlphaFold-generated structures. As a proof of concept, we examine the performance of SAnDReS-generated models in three case studies. For all three cases, our models outperformed classical scoring functions. Also, SAnDReS-generated models showed predictive performance close to or better than other machine-learning models such as KDEEP, CSM-lig, and ΔVinaRF20. SAnDReS 2.0 is available to download at https://github.com/azevedolab/sandres.

Chronic Behavioral and Neurochemical Effects of Four Novel N-Benzyl-2-phenylethylamine Derivatives Recently Identified as "Psychoactive" in Adult Zebrafish Screens.

Potently affecting human and animal brain and behavior, hallucinogenic drugs have recently emerged as potentially promising agents in psychopharmacotherapy. Complementing laboratory rodents, the zebrafish (Danio rerio) is a powerful model organism for screening neuroactive drugs, including hallucinogens. Here, we tested four novel N-benzyl-2-phenylethylamine (NBPEA) derivatives with 2,4- and 3,4-dimethoxy substitutions in the phenethylamine moiety and the -F, -Cl, and -OCF3 substitutions in the ortho position of the phenyl ring of the N-benzyl moiety (34H-NBF, 34H-NBCl, 24H-NBOMe(F), and 34H-NBOMe(F)), assessing their behavioral and neurochemical effects following chronic 14 day treatment in adult zebrafish. While the novel tank test behavioral data indicate anxiolytic-like effects of 24H-NBOMe(F) and 34H-NBOMe(F), neurochemical analyses reveal reduced brain norepinephrine by all four drugs, and (except 34H-NBCl) - reduced dopamine and serotonin levels. We also found reduced turnover rates for all three brain monoamines but unaltered levels of their respective metabolites. Collectively, these findings further our understanding of complex central behavioral and neurochemical effects of chronically administered novel NBPEAs and highlight the potential of zebrafish as a model for preclinical screening of small psychoactive molecules.

MetaTox 2.0: Estimating the Biological Activity Spectra of Drug-like Compounds Taking into Account Probable Biotransformations.

After the biotransformation of xenobiotics in the human body, the biological activity of the metabolites may differ from the activity of parent compounds. Therefore, to assess the overall biological activity of a drug-like compound, it is important to take into account its metabolites and their biological activity. We developed MetaTox 2.0-an updated version of the MetaTox web application that was able to predict the metabolites of xenobiotics. Innovations include estimating the biological activity profile of a compound and taking into account its metabolites. The estimation is based on the PASS (prediction of activity spectra for substances) algorithm and on the latest version of the training set covering over 1900 biological activities predicted with an average accuracy exceeding 0.97. Also, MetaTox 2.0 allows the search for similar substances among more than 2000 drugs with known metabolic networks, which were extracted from the ChEMBL, MetXBIODB, and DrugBank databases. MetaTox 2.0 is freely available on the web at https://www.way2drug.com/metatox.

Web Service for HIV Drug Resistance Prediction Based on Analysis of Amino Acid Substitutions in Main Drug Targets.

Predicting viral drug resistance is a significant medical concern. The importance of this problem stimulates the continuous development of experimental and new computational approaches. The use of computational approaches allows researchers to increase therapy effectiveness and reduce the time and expenses involved when the prescribed antiretroviral therapy is ineffective in the treatment of infection caused by the human immunodeficiency virus type 1 (HIV-1). We propose two machine learning methods and the appropriate models for predicting HIV drug resistance related to amino acid substitutions in HIV targets: (i) k-mers utilizing the random forest and the support vector machine algorithms of the scikit-learn library, and (ii) multi-n-grams using the Bayesian approach implemented in MultiPASSR software. Both multi-n-grams and k-mers were computed based on the amino acid sequences of HIV enzymes: reverse transcriptase and protease. The performance of the models was estimated by five-fold cross-validation. The resulting classification models have a relatively high reliability (minimum accuracy for the drugs is 0.82, maximum: 0.94) and were used to create a web application, HVR (HIV drug Resistance), for the prediction of HIV drug resistance to protease inhibitors and nucleoside and non-nucleoside reverse transcriptase inhibitors based on the analysis of the amino acid sequences of the appropriate HIV proteins from clinical samples.

Transcriptome-based analysis of human peripheral blood reveals regulators of immune response in different viral infections.

Introduction: There are difficulties in creating direct antiviral drugs for all viruses, including new, suddenly arising infections, such as COVID-19. Therefore, pathogenesis-directed therapy is often necessary to treat severe viral infections and comorbidities associated with them. Despite significant differences in the etiopathogenesis of viral diseases, in general, they are associated with significant dysfunction of the immune system. Study of common mechanisms of immune dysfunction caused by different viral infections can help develop novel therapeutic strategies to combat infections and associated comorbidities. Methods: To identify common mechanisms of immune functions disruption during infection by nine different viruses (cytomegalovirus, Ebstein-Barr virus, human T-cell leukemia virus type 1, Hepatitis B and C viruses, human immunodeficiency virus, Dengue virus, SARS-CoV, and SARS-CoV-2), we analyzed the corresponding transcription profiles from peripheral blood mononuclear cells (PBMC) using the originally developed pipeline that include transcriptome data collection, processing, normalization, analysis and search for master regulators of several viral infections. The ten datasets containing transcription data from patients infected by nine viruses and healthy people were obtained from Gene Expression Omnibus. The analysis of the data was performed by Genome Enhancer pipeline. Results: We revealed common pathways, cellular processes, and master regulators for studied viral infections. We found that all nine viral infections cause immune activation, exhaustion, cell proliferation disruption, and increased susceptibility to apoptosis. Using network analysis, we identified PBMC receptors, representing proteins at the top of signaling pathways that may be responsible for the observed transcriptional changes and maintain the current functional state of cells. Discussion: The identified relationships between some of them and virus-induced alteration of immune functions are new and have not been found earlier, e.g., receptors for autocrine motility factor, insulin, prolactin, angiotensin II, and immunoglobulin epsilon. Modulation of the identified receptors can be investigated as one of therapeutic strategies for the treatment of severe viral infections.

HGMMX: Host Gut Microbiota Metabolism Xenobiotics Database.

The metagenome of bacteria colonizing the human intestine is a set of genes that is almost 150 times greater than the set of host genes. Some of these genes encode enzymes whose functioning significantly expands the number of potential pathways for xenobiotic metabolism. The resulting metabolites can exhibit activity different from that of the parent compound. This can decrease the efficacy of pharmacotherapy as well as induce undesirable and potentially life-threatening side effects. Thus, analysis of the biotransformation of small drug-like compounds mediated by the gut microbiota is an important step in the development of new pharmaceutical agents and repurposing of the approved drugs. In vitro research, the interaction of drug-like compounds with the gut microbiota is a multistep and time-consuming process. Systematic testing of large sets of chemical structures is associated with a number of challenges, including the lack of standardized techniques and significant financial costs to identify the structure of the final metabolites. Estimation of the compounds' ability to be biotransformed by the gut microbiota and prediction of the structures of their metabolites are possible in silico. However, the development of computational approaches is limited by the lack of information about chemical structures metabolized by microbiota enzymes. The aim of this study is to create a database containing information on the metabolism of drug-like compounds by the gut microbiota. We created the data set containing information about 368 structures metabolized and 310 structures not metabolized by the human gut microbiota. The HGMMX database is freely available at https://www.way2drug.com/hgmmx. The information presented will be useful in the development of computational approaches for analyzing the impact of the human microbiota on metabolism of drug-like molecules.

Towards Novel Potential Molecular Targets for Antidepressant and Antipsychotic Pharmacotherapies.

Depression and schizophrenia are two highly prevalent and severely debilitating neuropsychiatric disorders. Both conventional antidepressant and antipsychotic pharmacotherapies are often inefficient clinically, causing multiple side effects and serious patient compliance problems. Collectively, this calls for the development of novel drug targets for treating depressed and schizophrenic patients. Here, we discuss recent translational advances, research tools and approaches, aiming to facilitate innovative drug discovery in this field. Providing a comprehensive overview of current antidepressants and antipsychotic drugs, we also outline potential novel molecular targets for treating depression and schizophrenia. We also critically evaluate multiple translational challenges and summarize various open questions, in order to foster further integrative cross-discipline research into antidepressant and antipsychotic drug development.

Phyto4Health: Database of Phytocomponents from Russian Pharmacopoeia Plants.

Medicinal plants growing in Russia are a rich source of biologically active compounds. However, the evaluation of the hidden pharmacological potential of these compounds by in silico methods is complicated by the lack of specialized databases. We have created a database of 3128 phytocomponents from 268 medical plants included in the Russian Pharmacopoeia. The information about the compounds was supplemented with their physical-chemical characteristics and biological activity profiles estimated using the PASS software. Comparison with phytocomponents of medicinal plants from five other countries showed that the similarity of phytocomponents in our database is rather small. The uniqueness of the contents significantly enriches and provides easy access to the necessary information. The Phyto4Health data are freely available at http://www.way2drug.com/p4h/.

Exploring Scoring Function Space: Developing Computational Models for Drug Discovery.

BACKGROUND: The idea of scoring function space established a systems-level approach to address the development of models to predict the affinity of drug molecules by those interested in drug discovery. OBJECTIVE: Our goal here is to review the concept of scoring function space and how to explore it to develop machine learning models to address protein-ligand binding affinity. METHOD: We searched the articles available in PubMed related to the scoring function space. We also utilized crystallographic structures found in the protein data bank (PDB) to represent the protein space. RESULTS: The application of systems-level approaches to address receptor-drug interactions allows us to have a holistic view of the process of drug discovery. The scoring function space adds flexibility to the process since it makes it possible to see drug discovery as a relationship involving mathematical spaces. CONCLUSION: The application of the concept of scoring function space has provided us with an integrated view of drug discovery methods. This concept is useful during drug discovery, where we see the process as a computational search of the scoring function space to find an adequate model to predict receptor-drug binding affinity.

Identification of Molecular Mechanisms Involved in Viral Infection Progression Based on Text Mining: Case Study for HIV Infection.

Viruses cause various infections that may affect human lifestyle for durations ranging from several days to for many years. Although preventative and therapeutic remedies are available for many viruses, they may still have a profound impact on human life. The human immunodeficiency virus type 1 is the most common cause of HIV infection, which represents one of the most dangerous and complex diseases since it affects the immune system and causes its disruption, leading to secondary complications and negatively influencing health-related quality of life. While highly active antiretroviral therapy may decrease the viral load and the velocity of HIV infection progression, some individual peculiarities may affect viral load control or the progression of T-cell malfunction induced by HIV. Our study is aimed at the text-based identification of molecular mechanisms that may be involved in viral infection progression, using HIV as a case study. Specifically, we identified human proteins and genes which commonly occurred, overexpressed or underexpressed, in the collections of publications relevant to (i) HIV infection progression and (ii) acute and chronic stages of HIV infection. Then, we considered biological processes that are controlled by the identified protein and genes. We verified the impact of the identified molecules in the associated clinical study.

CLC-Pred 2.0: A Freely Available Web Application for In Silico Prediction of Human Cell Line Cytotoxicity and Molecular Mechanisms of Action for Druglike Compounds.

In vitro cell-line cytotoxicity is widely used in the experimental studies of potential antineoplastic agents and evaluation of safety in drug discovery. In silico estimation of cytotoxicity against hundreds of tumor cell lines and dozens of normal cell lines considerably reduces the time and costs of drug development and the assessment of new pharmaceutical agent perspectives. In 2018, we developed the first freely available web application (CLC-Pred) for the qualitative prediction of cytotoxicity against 278 tumor and 27 normal cell lines based on structural formulas of 59,882 compounds. Here, we present a new version of this web application: CLC-Pred 2.0. It also employs the PASS (Prediction of Activity Spectra for Substance) approach based on substructural atom centric MNA descriptors and a Bayesian algorithm. CLC-Pred 2.0 provides three types of qualitative prediction: (1) cytotoxicity against 391 tumor and 47 normal human cell lines based on ChEMBL and PubChem data (128,545 structures) with a mean accuracy of prediction (AUC), calculated by the leave-one-out (LOO CV) and the 20-fold cross-validation (20F CV) procedures, of 0.925 and 0.923, respectively; (2) cytotoxicity against an NCI60 tumor cell-line panel based on the Developmental Therapeutics Program's NCI60 data (22,726 structures) with different thresholds of IG50 data (100, 10 and 1 nM) and a mean accuracy of prediction from 0.870 to 0.945 (LOO CV) and from 0.869 to 0.942 (20F CV), respectively; (3) 2170 molecular mechanisms of actions based on ChEMBL and PubChem data (656,011 structures) with a mean accuracy of prediction 0.979 (LOO CV) and 0.978 (20F CV). Therefore, CLC-Pred 2.0 is a significant extension of the capabilities of the initial web application.

Blood Plasma Proteome: A Meta-Analysis of the Results of Protein Quantification in Human Blood by Targeted Mass Spectrometry.

A meta-analysis of the results of targeted quantitative screening of human blood plasma was performed to generate a reference standard kit that can be used for health analytics. The panel included 53 of the 296 proteins that form a "stable" part of the proteome of a healthy individual; these proteins were found in at least 70% of samples and were characterized by an interindividual coefficient of variation <40%. The concentration range of the selected proteins was 10−10−10−3 M and enrichment analysis revealed their association with rare familial diseases. The concentration of ceruloplasmin was reduced by approximately three orders of magnitude in patients with neurological disorders compared to healthy volunteers, and those of gelsolin isoform 1 and complement factor H were abruptly reduced in patients with lung adenocarcinoma. Absolute quantitative data of the individual proteome of a healthy and diseased individual can be used as the basis for personalized medicine and health monitoring. Storage over time allows us to identify individual biomarkers in the molecular landscape and prevent pathological conditions.

New Positive TRPC6 Modulator Penetrates Blood-Brain Barrier, Eliminates Synaptic Deficiency and Restores Memory Deficit in 5xFAD Mice.

Synapse loss in the brain of Alzheimer's disease patients correlates with cognitive dysfunctions. Drugs that limit synaptic loss are promising pharmacological agents. The transient receptor potential cation channel, subfamily C, member 6 (TRPC6) regulates the formation of an excitatory synapse. Positive regulation of TRPC6 results in increased synapse formation and enhances learning and memory in animal models. The novel selective TRPC6 agonist, 3-(3-,4-Dihydro-6,7-dimethoxy-3,3-dimethyl-1-isoquinolinyl)-2H-1-benzopyran-2-one, has recently been identified. Here we present in silico, in vitro, ex vivo, pharmacokinetic and in vivo studies of this compound. We demonstrate that it binds to the extracellular agonist binding site of the human TRPC6, protects hippocampal mushroom spines from amyloid toxicity in vitro, efficiently recovers synaptic plasticity in 5xFAD brain slices, penetrates the blood-brain barrier and recovers cognitive deficits in 5xFAD mice. We suggest that C20 might be recognized as the novel TRPC6-selective drug suitable to treat synaptic deficiency in Alzheimer's disease-affected hippocampal neurons.

Inhibitory Effects of 7-Methylguanine and Its Metabolite 8-Hydroxy-7-Methylguanine on Human Poly(ADP-Ribose) Polymerase 1.

Previously, we have found that a nucleic acid metabolite, 7-methylguanine (7mGua), produced in the body can have an inhibitory effect on the poly(ADP-ribose) polymerase 1 (PARP1) enzyme, an important pharmacological target in anticancer therapy. In this work, using an original method of analysis of PARP1 activity based on monitoring fluorescence anisotropy, we studied inhibitory properties of 7mGua and its metabolite, 8-hydroxy-7-methylguanine (8h7mGua). Both compounds inhibited PARP1 enzymatic activity in a dose-dependent manner, however, 8h7mGua was shown to be a stronger inhibitor. The IC50 values for 8h7mGua at different concentrations of the NAD+ substrate were found to be 4 times lower, on average, than those for 7mGua. The more efficient binding of 8h7mGua in the PARP1 active site is explained by the presence of an additional hydrogen bond with the Glu988 catalytic residue. Experimental and computational studies did not reveal the effect of 7mGua and 8h7mGua on the activity of other DNA repair enzymes, indicating selectivity of their inhibitory action.

Computational Prediction of Inhibitors and Inducers of the Major Isoforms of Cytochrome P450.

Human cytochrome P450 enzymes (CYPs) are heme-containing monooxygenases. This superfamily of drug-metabolizing enzymes is responsible for the metabolism of most drugs and other xenobiotics. The inhibition of CYPs may lead to drug-drug interactions and impair the biotransformation of drugs. CYP inducers may decrease the bioavailability and increase the clearance of drugs. Based on the freely available databases ChEMBL and PubChem, we have collected over 70,000 records containing the structures of inhibitors and inducers together with the IC50 values for the inhibitors of the five major human CYPs: 1A2, 3A4, 2D6, 2C9, and 2C19. Based on the collected data, we developed (Q)SAR models for predicting inhibitors and inducers of these CYPs using GUSAR and PASS software. The developed (Q)SAR models could be applied for assessment of the interaction of novel drug-like substances with the major human CYPs. The created (Q)SAR models demonstrated reasonable accuracy of prediction. They have been implemented in the web application P450-Analyzer that is freely available via the Internet.

Novel mitochondria-targeting compounds selectively kill human leukemia cells.

Acute myeloid leukemia (AML) is a heterogeneous group of aggressive hematological malignancies commonly associated with treatment resistance, high risk of relapse, and mitochondrial dysregulation. We identified six mitochondria-affecting compounds (PS compounds) that exhibit selective cytotoxicity against AML cells in vitro. Structure-activity relationship studies identified six analogs from two original scaffolds that had over an order of magnitude difference between LD50 in AML and healthy peripheral blood mononuclear cells. Mechanistically, all hit compounds reduced ATP and selectively impaired both basal and ATP-linked oxygen consumption in leukemic cells. Compounds derived from PS127 significantly upregulated production of reactive oxygen species (ROS) in AML cells and triggered ferroptotic, necroptotic, and/or apoptotic cell death in AML cell lines and refractory/relapsed AML primary samples. These compounds exhibited synergy with several anti-leukemia agents in AML, acute lymphoblastic leukemia (ALL), or chronic myelogenous leukemia (CML). Pilot in vivo efficacy studies indicate anti-leukemic efficacy in a MOLM14/GFP/LUC xenograft model, including extended survival in mice injected with leukemic cells pre-treated with PS127B or PS127E and in mice treated with PS127E at a dose of 5 mg/kg. These compounds are promising leads for development of future combinatorial therapeutic approaches for mitochondria-driven hematologic malignancies such as AML, ALL, and CML.

The method predicting interaction between protein targets and small-molecular ligands with the wide applicability domain.

Prediction of protein-ligand interaction is necessary for drug design, gene regulatory networks investigation, and chemical probes detection. The existing methods commonly demonstrate high prediction accuracy for the particular groups of protein and their ligands. We developed an approach suited for the wider applicability and tested it on three dataset types significantly differing by protein homology. The study included three typical scenarios of assessing the target-ligand interaction: 1st - predicting protein targets by ligand structures' comparisons; 2nd - predicting ligands by target sequences' comparisons; 3rd - predicting both the uncharacterized targets and ligands with the fuzzy coefficients based on ligand comparisons. The 1st scenario implemented showed a high prediction accuracy of 0.96-0.99, providing fuzzy coefficients of target-ligand interactions in the 3rd scenario. Testing by 2nd scenario displayed the accuracy of 0.97-0.99 for predicting within the particular protein families, sets non-ordered by protein homology, and accuracy higher than 0.90 for most HIV sets, each presenting the close mutant proteins differing by point substitutions. The 3rd scenario displayed that fuzzy classification can reveal reasonable accuracy 0.86-0.94 at simulated data incompleteness. Thus, our approach provides high prediction accuracy with the wide applicability domain, including data differing in heterogeneity and completeness.

HIV-1 drug resistance profiling using amino acid sequence space cartography.

MOTIVATION: Human immunodeficiency virus (HIV) drug resistance is a global healthcare issue. The emergence of drug resistance influenced the efficacy of treatment regimens, thus stressing the importance of treatment adaptation. Computational methods predicting the drug resistance profile from genomic data of HIV isolates are advantageous for monitoring drug resistance in patients. However, existing computational methods for drug resistance prediction are either not suitable for emerging HIV strains with complex mutational patterns or lack interpretability, which is of paramount importance in clinical practice. The approach reported here overcomes these limitations and combines high accuracy of predictions and interpretability of the models. RESULTS: In this work, a new methodology based on generative topographic mapping (GTM) for biological sequence space representation and quantitative genotype-phenotype relationships prediction purposes was introduced. The GTM-based resistance landscapes allowed us to predict the resistance of HIV strains based on sequencing and drug resistance data for three viral proteins [integrase (IN), protease (PR) and reverse transcriptase (RT)] from Stanford HIV drug resistance database. The average balanced accuracy for PR inhibitors was 0.89 ± 0.01, for IN inhibitors 0.85 ± 0.01, for non-nucleoside RT inhibitors 0.73 ± 0.01 and for nucleoside RT inhibitors 0.84 ± 0.01. We have demonstrated in several case studies that GTM-based resistance landscapes are useful for visualization and analysis of sequence space as well as for treatment optimization purposes. Here, GTMs were applied for the in-depth analysis of the relationships between mutation pattern and drug resistance using mutation landscapes. This allowed us to predict retrospectively the importance of the presence of particular mutations (e.g. V32I, L10F and L33F in HIV PR) for the resistance development. This study highlights some perspectives of GTM applications in clinical informatics and particularly in the field of sequence space exploration. AVAILABILITY AND IMPLEMENTATION: https://github.com/karinapikalyova/ISIDASeq. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

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.