QSAR Methods Research Articles

Novel α-glucosidase Inhibitors Designed as Type 2 Diabetes Drugs by QSAR, Molecular Docking and Molecular Dynamics Simulation Methods.

Diabetes mellitus is a globally prevalent disease of significant concern. Alpha-glucosidase has emerged as a prominent target for the treatment of type 2 diabetes. In this study, 39 α-glucosidase inhibitors (AGIs) of tetrahydrobenzo[b]thiophene-2-ylurea derivatives to establish a stable and valid Topomer CoMFA model, with a cross-validation coefficient (q²) of 0.766 and a non-cross-validation coefficient (r²) of 0.960. Subsequently, the ZINC15 database was used to screen the fragments, based on which 13 novel inhibitor molecules with theoretically potentially high activity were designed. Molecular docking and molecular dynamics simulations to understand the binding status of the inhibitor molecules to the target proteins showed that amino acids ASP215, GLN279 and ARG442 may form hydrogen bonds with the ligands and therefore enhance the inhibitory effect of the small molecules. Additionally, MM/PBSA calculations suggest that the newly designed molecules demonstrate more stable binding modes, and these newly designed molecules showed good ADMET properties with potential as AGIs. The findings would provide valuable guidance and a theoretical foundation for the design and development of novel AGIs.

Transformers enable accurate prediction of acute and chronic chemical toxicity in aquatic organisms.

Environmental hazard assessments are reliant on toxicity data that cover multiple organism groups. Generating experimental toxicity data is, however, resource-intensive and time-consuming. Computational methods are fast and cost-efficient alternatives, but the low accuracy and narrow applicability domains have made their adaptation slow. Here, we present a AI-based model for predicting chemical toxicity. The model uses transformers to capture toxicity-specific features directly from the chemical structures and deep neural networks to predict effect concentrations. The model showed high predictive performance for all tested organism groups-algae, aquatic invertebrates and fish-and has, in comparison to commonly used QSAR methods, a larger applicability domain and a considerably lower error. When the model was trained on data with multiple effect concentrations (EC50/EC10), the performance was further improved. We conclude that deep learning and transformers have the potential to markedly advance computational prediction of chemical toxicity.

Prediction of selected properties of aflatoxin molecules by the QSAR method

Abstract Aflatoxins are naturally occurring compounds produced by fungi, mainly of Aspergillus species. All aflatoxins are proved to cause acute toxicity to human health, some even causing chronic diseases such as cancer. These molecules clearly proved that even natural molecules can be directly related to cancer and the formation of tumours. In general, aflatoxins can be characterised as organic compounds; B1, G1, B2, and G2 are produced directly by fungi and these four are subsequently metabolised in biological systems of microbes, animals or humans into other forms, such as M1, EB1, AFL etc. This manuscript provides a brief overview of 14 aflatoxins, their molecular structure and its possible relationship to aflatoxins biological activity. This information, in combination with additional calculations, offers the possibility to investigate the mentioned compounds and their properties using the QSAR approach.

Construction of a floating photothermal-assisted photocatalytic system with a three-dimensional hollow porous network structure

Solar energy is the inevitable choice to achieve the low-carbon, green, and circular development of society, and photocatalysis technology is one of the shining pearls. To make full use of the solar spectrum and solve the shortcomings of the recovery difficulty of powdery materials and the loss of activity due to the influence of the external environment, it is possible to construct floating materials using melamine sponges to recover photocatalytic materials quickly. At the same time, floating materials can absorb oxygen in the air for the generation of active groups, effectively solving the problem of less O2 in the water. The carbon-based materials have excellent light absorption properties, high thermal conductivity, and excellent photothermal conversion efficiency and are ideal for constructing floating photothermal photocatalytic systems. As an example, we combined a cheap melamine sponge with urea, prepared a hollow porous network structure g-C3N4 (HPNCN) with a high specific surface area by direct thermal shrinkage method, and then attached the CoO to its surface by hydrothermal method to form a heterojunction with a suitable band gap. Various characterization tests verified the photothermal-photocatalytic properties. Among them, 30% CoO/HPNCN has the best photocatalytic degradation effect on tetracycline (TC), and the removal rate is 88.1%. After five cycles, the removal rate is only 5% lower than the initial, indicating that it has good stability and recyclability. We conducted an active ingredient capture experiment, ESR, and LC-MS analysis to clarify the intermediates and reaction mechanism of TC photocatalytic degradation. On this basis, the ECOSAR program and QSAR method were used to analyze the environmental toxicity of TC and its intermediate products. These results provide a broad prospect for the potential application of the floating photothermal-photocatalysis system in antibiotic pollution control and its application in other fields.

Potential Health Risks of Chloroacetanilide Herbicides: An In Silico Analysis

The extensive use of herbicidal products in agriculture and forestry has raised concerns over potential adverse effects on human health and the environment. Chloroacetanilide herbicides are a group of synthetic chemicals used to control weeds in agriculture and forestry. However, so[me of their members have been characterized as possible carcinogens. The genotoxicity and carcinogenicity of two chloroacetanilide herbicides, delachlor and xylachlor, are discussed. This article proposes to use tools to predict their potential toxicities based on their chemical structure. Four software tools, Vega Hub, Toxtree, Lazar, and TEST, are used to predict the potential genotoxic and carcinogenic effects of the herbicides. Vega Hub uses QSAR models, Toxtree uses a decision tree approach, Lazar uses data mining algorithms, and TEST uses QSAR methods to estimate toxicity. The canonical Simplified Molecular Input Line Entry Specification (SMILES) systems of delachlor and xylachlor are entered into each software tool to create a prediction. The study found that delachlor and xylachlor is a class 3 highly toxic compounds with potential mutagenic and carcinogenic effects based on Toxtree and Vega Hub. Meanwhile, Lazar and TEST predicted that delachlor and xylachlor are unlikely to be mutagenic. This study to determine the toxicity of the herbicides delachlor and xylachlor has shown that the possible effects of these herbicides on health and the environment need to be further investigated. The results provide valuable insights into chloroacetanilide herbicide toxicity and help develop safer, more environmentally friendly alternatives.

Bioactive Steroids Bearing Oxirane Ring.

This review explores the biological activity and structural diversity of steroids and related isoprenoid lipids, with a particular focus on compounds containing an oxirane ring. These natural compounds are derived from fungi, fungal endophytes, as well as extracts of plants, algae, and marine invertebrates. To evaluate their biological activity, an extensive examination of refereed literature sources was conducted, including in vivo and in vitro studies and the utilization of the QSAR method. Notable properties observed among these compounds include strong anti-inflammatory, antineoplastic, antiproliferative, anti-hypercholesterolemic, antiparkinsonian, diuretic, anti-eczematic, anti-psoriatic, and various other activities. Throughout this review, 3D graphs illustrating the activity of individual steroids are presented, accompanied by images of selected terrestrial or marine organisms. Furthermore, this review provides explanations for specific types of biological activity associated with these compounds. The data presented in this review are of scientific interest to the academic community and carry practical implications in the fields of pharmacology and medicine. By analyzing the biological activity and structural diversity of steroids and related isoprenoid lipids, this review offers valuable insights that contribute to both theoretical understanding and applied research. This review draws upon data from various authors to compile information on the biological activity of natural steroids containing an oxirane ring.

Biological Activity and Structural Diversity of Steroids Containing Aromatic Rings, Phosphate Groups, or Halogen Atoms.

This review delves into the investigation of the biological activity and structural diversity of steroids and related isoprenoid lipids. The study encompasses various natural compounds, such as steroids with aromatic ring(s), steroid phosphate esters derived from marine invertebrates, and steroids incorporating halogen atoms (I, Br, or Cl). These compounds are either produced by fungi or fungal endophytes or found in extracts of plants, algae, or marine invertebrates. To assess the biological activity of these natural compounds, an extensive examination of referenced literature sources was conducted. The evaluation encompassed in vivo and in vitro studies, as well as the utilization of the QSAR method. Numerous compounds exhibited notable properties such as strong anti-inflammatory, anti-neoplastic, anti-proliferative, anti-hypercholesterolemic, anti-Parkinsonian, diuretic, anti-eczematic, anti-psoriatic, and various other activities. Throughout the review, 3D graphs illustrating the activity of individual steroids are presented alongside images of selected terrestrial or marine organisms. Additionally, the review provides explanations for specific types of biological activity associated with these compounds. The data presented in this review hold scientific interest for academic science as well as practical implications in the fields of pharmacology and practical medicine. The analysis of the biological activity and structural diversity of steroids and related isoprenoid lipids provides valuable insights that can contribute to advancements in both theoretical understanding and applied research.

INVESTIGATION OF TOXICOLOGICAL PROPERTIES OF SOME AZO DYES BY OECD QSAR METHOD

In this study, the toxicological properties of some of the aromatic amine raw materials commonly used in textile dyes and the dyestuffs produced from these raw materials were examined with the OECD QSAR method and their toxicological effects on human health were investigated. In the literature review conducted with the relevant structures, it was seen that no study set was found in terms of toxicology, considering the comparison of raw materials and products. Studies have found that some aromatic amine raw materials and azo dyes produced from these raw materials are generally moderately irritating. While the related raw materials generally have eye irritant properties, the products do not have any irritating properties. It can be said that 12 aromatic amine raw materials and 13 azo dyes have skin sensitizing properties. The raw materials are 2,6-dibromo-4-nitro aniline, 6-methoxybenzothiazole-2-ylamine, 2-Bromo-4,6-dinotroaniline, and azo dyes, which are Disperse Brown 27-1, Disperse Brown 19 and Disperse Blue 291, are mutagenic.

An Insightful Evaluation of Evodiamine Analogs Effect as DNA Topoisomerase I Inhibitors Using QSAR Method

An Insightful Evaluation of Evodiamine Analogs Effect as DNA Topoisomerase I Inhibitors Using QSAR Method

Exploring QSAR models for activity-cliff prediction

Introduction and methodologyPairs of similar compounds that only differ by a small structural modification but exhibit a large difference in their binding affinity for a given target are known as activity cliffs (ACs). It has been hypothesised that QSAR models struggle to predict ACs and that ACs thus form a major source of prediction error. However, the AC-prediction power of modern QSAR methods and its quantitative relationship to general QSAR-prediction performance is still underexplored. We systematically construct nine distinct QSAR models by combining three molecular representation methods (extended-connectivity fingerprints, physicochemical-descriptor vectors and graph isomorphism networks) with three regression techniques (random forests, k-nearest neighbours and multilayer perceptrons); we then use each resulting model to classify pairs of similar compounds as ACs or non-ACs and to predict the activities of individual molecules in three case studies: dopamine receptor D2, factor Xa, and SARS-CoV-2 main protease.Results and conclusionsOur results provide strong support for the hypothesis that indeed QSAR models frequently fail to predict ACs. We observe low AC-sensitivity amongst the evaluated models when the activities of both compounds are unknown, but a substantial increase in AC-sensitivity when the actual activity of one of the compounds is given. Graph isomorphism features are found to be competitive with or superior to classical molecular representations for AC-classification and can thus be employed as baseline AC-prediction models or simple compound-optimisation tools. For general QSAR-prediction, however, extended-connectivity fingerprints still consistently deliver the best performance amongs the tested input representations. A potential future pathway to improve QSAR-modelling performance might be the development of techniques to increase AC-sensitivity.Graphical

APPLICATION OF MATHEMATICAL MODELING AND PHYSICO-CHEMICAL ANALYSIS METHODS IN THE PREDICTION OF BIOLOGICAL ACTIVITY AND QUALITY CONTROL OF GOSSYPOL DERIVATIVES

Objective: The purpose of this work was to evaluate in silico biological activity profiles of real and virtual molecular structures of gossypol derivatives and to develop methods of Physico-chemical analysis to control their quality. Methods: Substance of gossypol-acetic acid (GAA) and 14 virtual derivatives; PASS and ChemicDescript QSAR methods; low angle and dynamic laser light scattering (LALLS, DLS) methods; IR Spectroscopy–Cary 630 Fourier Transform IR Spectrometer, UV spectrometry–Cary-60 spectrophotometer, Optical microscopy (Altami BIO 2 microscope); Spirotox method for a sample’s biological activity. Results: A distance-based topological Balaban index (J) was successfully selected by ChemicDescript analysis; the Pa meaning by PASS Online program showed maximum (from 0.8 to 0.9) variations of antitumor and antiandrogenic and minimum of antiviral activities of GAA derivatives (Pa<0.5) despite the existing literature data. Microscopy and DLS methods demonstrated the values of high powder dispersion d=0.8 nm and weak stability of colloidal particles =-0.9 mV. According to UV data =42.4±0.8 (100 ml·g-1·cm-1) at λmax=380 nm. The LALLS method determined the GAA dissolution rate constant in ethanol: k=0.041±0.004 s-1. The calculated activation energy values of cell biosensor death process in 1 mmol solution of GAA in N,N-DMF: °bsEa=174.36±0.45 kJ·mol-1 in comparison with the solvent medium: °bsEa=213±1.55 kJ·mol-1 Conclusion: The developed approach of chemometric, laser and biotesting methods can be used for the identification of biologically active properties, as well as for qualitative analysis within the development of the standard for the pharmaceutical substance of natural polyphenols.

Rationally designed Ta3N5/BiOCl S-scheme heterojunction with oxygen vacancies for elimination of tetracycline antibiotic and Cr(VI): Performance, toxicity evaluation and mechanism insight

S-scheme heterojunction photocatalysts have been the “stars” in the field of photocatalysis. Herein, a novel S-scheme heterojunction of Ta3N5/BiOCl with oxygen vacancies (OVs) was fabricated via a facile method. The charge separation and transport mechanism of this Ta3N5/BiOCl S-scheme heterojunction was verified by the analyses of band energy structures, active species, photoelectric behaviors and DFT theoretical calculation. Compared with Ta3N5 and BiOCl, the Ta3N5/BiOCl unveils substantially upgraded photocatalytic property under visible light, and the photocatalytic efficiency for removal of tetracycline (TC) and hexavalent chromium (Cr(VI)) reaches 89.6% and 91.6%, respectively. The substantial enhancement of the photocatalytic activity is attributed to the synergistic effect of the S-scheme hetero-structure and oxygen vacancies, which improves the visible-light absorption, while promoting the spatial separation of charge carriers with the optimum redox capacity, thereby boosting the production of active species for catalytic reactions. The TC degradation pathway is deduced and the toxicity evolution of TC is appraised using the QSAR method. In a nutshell, this work gives a deep understanding of the photocatalytic mechanism based on Ta3N5/BiOCl as well as presents a newfangled thought for developing highly efficient S-scheme heterojunction photocatalysts for water decontamination.

Surface atom rearrangement on carbon nitride for enhanced photocatalysis degradation of antibiotics under visible light

Recently, antibiotic pollution has aroused great concern due to its toxicity to living organisms and the risk of drug-resistance gene generation in the ecosystem. Carbon nitride (CN) photodegradation towards antibiotics is a promising technology for water purification. However, CN possesses a low activity due to the fast recombination of electron-hole pairs and slow reaction kinetics. Herein, we report the fabrication of nitrogen-vacancy and oxygen-substitution modified carbon nitride ((Nv, Os)–CN) with a rearranged surface via a low-temperature thermal oxidation strategy. The (Nv, Os)–CN can degrade tetracycline (TC), ciprofloxacin (CIP) and sulfadiazine (SZO) than CN under visible light irradiation efficiently. The first-order kinetic constant of (Nv, Os)–CN for TC, CIP and SZO was 2.1, 2.24 and 1.38 folds higher than those of CN, respectively. This high performance benefits from the rearranged surface atoms by forming an active surface with rich Nv and Os, thus promoting interfacial charge transfer and building robust water-(Nv, Os)–CN interface. In addition, the possible photodegraded pathways, resulting intermediates and risks of TC photodegradation were analyzed by the LC-MS and QSAR methods. Our proposed low-temperature thermal oxidation strategy to achieve surface atom rearrangement endows CN with an enhanced photoactivity toward antibiotics degradation.

A novel approach to predict the comprehensive EROD potency: Mechanism-based curve fitting of CYP1A1 activity by PAHs

Cytochrome P450 1A1 (CYP1A1) plays critical roles in polycyclic aromatic hydrocarbon (PAH) toxicity, including DNA adduction and ROS generation. Therefore, CYP1A1 activity quantified by the 7-ethoxyresorufin-O-deethylase (EROD) assay (named EROD potency) has been considered a typical biomarker of PAH exposure and toxicity. The EROD dose–response curve always presents a biphasic style, increasing at low concentrations and decreasing at high concentrations of PAHs, but relative effect potency (REP) commonly used in PAH risk assessment is only involved in the increasing phase. In this study, a full bell-shaped EROD curve fitting formula Eq. (1) was obtained by considering both CYP1A1 mRNA induction and enzyme inhibition to completely assess the EROD potency of PAHs. Correspondingly, in silico models of QSAR and docking methods successfully predicted the full EROD curves of PAHs, and the structure-activity relationship indicated that PAHs with heavy molecular weight and large diameter showed stronger EROD potency. Further EROD potency with predicted curve parameters (EC50,ind and area index) was confirmed by the reported REP (R2 = 0.697–0.977) and experimental data from human and mouse cells (R2 = 0.700–0.804). This study provides a novel curve fitting for the EROD dose–response relationship and a prediction model for PAH EROD potency.

An Inverse QSAR Method Based on Linear Regression and Integer Programming

Drug design is one of the important applications of biological science. Extensive studies have been done on computer-aided drug design based on inverse quantitative structure activity relationship (inverse QSAR), which is to infer chemical compounds from given chemical activities and constraints. However, exact or optimal solutions are not guaranteed in most of the existing methods. Recently a novel framework based on artificial neural networks (ANNs) and mixed integer linear programming (MILP) has been proposed for designing chemical structures. This framework consists of two phases: an ANN is used to construct a prediction function, and then an MILP formulated on the trained ANN and a graph search algorithm are used to infer desired chemical structures. In this paper, we use linear regression instead of ANNs to construct a prediction function. For this, we derive a novel MILP formulation that simulates the computation process of a prediction function by linear regression. For the first phase, we performed computational experiments using 18 chemical properties, and the proposed method achieved good prediction accuracy for a relatively large number of properties, in comparison with ANNs in our previous work. For the second phase, we performed computational experiments on five chemical properties, and the method could infer chemical structures with around up to 50 non-hydrogen atoms. Combination of linear regression and integer programming is a potentially useful approach to computational molecular design.

Functionalization of magnetic chitosan microparticles – Comparison of trione and trithione grafting for enhanced silver sorption and application to metal recovery from waste X-ray photographic films

Magnetic chitosan microparticles (MC) are successfully functionalized by grafting two pyrimidine derivatives (bearing either trione or trithione groups for MC-PYO and MC-PYS, respectively). The sorbents are characterized using SEM, TEM, BET, TGA, FTIR, titration, and elemental analysis. The study focuses on the comparison of the differential effects of functionalized groups on the sorption of silver. FTIR analysis shows the contributions of N-based, carbonyl, hydroxyl for silver binding onto MC-PYO, completed by sulfur contribution in the case of MC-PYS. Maximum sorption capacities at pH 6 reach 1.9 and 2.3 mmol Ag g−1 for MC-PYO and MC-PYS, respectively. The Langmuir and the Sips equations fit sorption isotherms. The reactive groups affect the thermodynamic characteristics: endothermic for MC-PYS against exothermic for MC-PYO. The affinity of S-based soft ligand in MC-PYS for soft metals makes the sorbent selective for Ag(I). On the opposite hand, for MC-PYO the O-based functional groups readily bind hard metals (HSAB principle), meaning less selective for silver recovery (from multi-component solutions). QSAR method (quantitative structure-activity relationships) allows correlating preferentially sorption properties with the covalent index. Nitric acid solutions (0.3 M) are highly efficient for total desorption of silver from metal-loaded sorbents, but soft enough to maintain remarkable sorption and desorption performances for at least five cycles. The sorbents are successfully applied to Ag(I) recovery from acidic leachate of waste photographic films: metal removal is enhanced at pH 5.2; MC-PYS is more efficient and selective than MC-PYO at pH 2.2.

Deriving the predicted no effect concentrations of 35 pesticides by the QSAR-SSD method

The widespread use of pesticides results in their frequent detection in water bodies and other environmental media. Pesticide residues may cause certain risks to the environment and human health, and reliable predicted no effect concentrations (PNEC) must be obtained when assessing environmental risks. Species sensitivity distribution (SSD) is an important method for the derivation of chemical PNECs. Construction of the SSD model requires sufficient toxicity data to various species including at least eight families in three phyla, suitable nonlinear fitting functions and assessment factors (AFs) with certain uncertainty. However, most chemicals could not collect sufficient species toxicity data, while some chemicals had sufficient species toxicity data but could not find suitable fitting functions, thus hindering the construction of effective SSD models. To this end, the established QSAR models were applied to predict toxicity of chemicals to specific species to fill in the toxicity data gaps required for SSD and selecting multiple nonlinear functions to optimize the SSD model. Combined with QSAR and SSD methods, a new method of PNEC derivation was developed and successfully applied to the derivation of PNEC for 35 pesticides. Three QSAR models were used to predict the toxicities of six pesticides with few toxicity data. Nine two-parameter nonlinear functions were used to fit the toxicity-cumulative probability data one by one to determine the optimal SSD models. The hazardous concentrations at the cumulative probability of 5% and 10%, i. e, HC5 and HC10, respectively, were calculated by the optimal SSD model. The assessment factor used to determine the PNEC of the chemical based on the HC10 was derived from the quantitative correlation between HC10 and HC5 of pesticides found in this study. When the toxicity data are insufficient, it may be more appropriate to calculate the PNECs of chemicals using HC10 than using HC5.

A Review on Procedure of QSAR Assessment in Organic Compounds As a Measure of Antioxidant Potentiality

Chemical and biological properties of substances may be inferred from their more fundamental physical, chemical, and biological characteristics using QSAR models. An insilico model may be built using QSAR to anticipate the activity of novel molecules before they are synthesised, allowing the author to establish a quantifiable link between structure and behaviour. QSAR is a powerful tool. Although QSAR modelling is a computer area, medicinal chemists are the main users and ultimate assessors, especially when it comes to developing compounds with the necessary biological activity. Several studies were conducted in which medicinal chemists and cheminformaticians collaborated to discover new compounds with specific biological activity. This was done through the development of QSAR models and their use in virtual screening, followed by experimental verification. Despite the fact that QSAR methods have their own set of limitations, their use in molecular prediction and assessment has been effective due to a division of labour in which mathematical professionals ensured the greatest quality of models. The predictions also helped experimental chemists design and test compounds that were expected to be successful. This review is being developed and implemented to look into the development of the QSAR tool in the assessment of antioxidant potentiality for diverse organic chemicals found in our environment.

Coupling band structure and oxidation-reduction potential to expound photodegradation performance difference of biochar-derived dissolved black carbon for organic pollutants under light irradiation

Herein, the photodegradation performances difference of rice straw biochar-derived dissolved black carbon (DBC) for Tetracycline and Methylene Blue under visible light irradiation have been investigated. Tetracycline is easier degraded (degradation rate: 68%), followed by Methylene Blue (degradation rate: 14%). Singlet oxygen (1O2), superoxide radicals (O2−), holes (h+) and triplet DBC (3DBC*) are all make contribution for Tetracycline degradation by DBC, whereas just singlet oxygen (1O2), superoxide radicals (O2−) and 3DBC* are involved in the MB degradation by DBC. Singlet oxygen (1O2) maybe from the fulvic acid-like structure of DBC, while band structure of DBC can explain why superoxide radicals (O2−) and holes (h+) can be formed, whereas hydroxyl radicals (OH) cannot be formed. The oxidation-reduction potential results of Tetracycline and Methylene Blue suggests that Tetracycline is easier to be oxidized than Methylene Blue as well as Methylene Blue is easier to be reduced than Tetracycline. Furthermore, experimental and theoretical results support that DBC has good interaction with Tetracycline, but the interaction between DBC and Methylene Blue is very weak. This likely explain why holes (h+) is not detected for Methylene Blue degradation by DBC since Methylene Blue has not too much chance to meet holes (h+). TC photodegradation intermediates are less toxic than Tetracycline based on QSAR method. Two possible photodegradation path of Tetracycline by DBC are proposed according to HPLC-MS results.

RespiraTox – Development of a QSAR model to predict human respiratory irritants

Respiratory irritation is an important human health endpoint in chemical risk assessment. There are two established modes of action of respiratory irritation, 1) sensory irritation mediated by the interaction with sensory neurons, potentially stimulating trigeminal nerve, and 2) direct tissue irritation. The aim of our research was to, develop a QSAR method to predict human respiratory irritants, and to potentially reduce the reliance on animal testing for the identification of respiratory irritants. Compounds are classified as irritating based on combined evidence from different types of toxicological data, including inhalation studies with acute and repeated exposure.The curated project database comprised 1997 organic substances, 1553 being classified as irritating and 444 as non-irritating. A comparison of machine learning approaches, including Logistic Regression (LR), Random Forests (RFs), and Gradient Boosted Decision Trees (GBTs), showed, the best classification was obtained by GBTs. The LR model resulted in an area under the curve (AUC) of 0.65, while the optimal performance for both RFs and GBTs gives an AUC of 0.71. In addition to the classification and the information on the applicability domain, the web-based tool provides a list of structurally similar analogues together with their experimental data to facilitate expert review for read-across purposes.