Existing and Developing Approaches for QSAR Analysis of Mixtures

Quantitative structure-insecticidal activity of essential oils on the human head louse (Pediculus humanus capitis)

In the search for natural and non-toxic products alternatives to synthetic pesticides, the fumigant and repellent activities of 35 essential oils are predicted in the human head louse (Pediculus humanus capitis) through the Quantitative Structure-Activity Relationships (QSAR) theory. The number of constituents of essential oils with weight percentage composition greater than 1% varies from 1 to 15, encompassing up to 213 structurally diverse compounds in the entire dataset. The 27,976 structural descriptors used to characterizing these complex mixtures are calculated as linear combinations of non-conformational descriptors for the components. This approach is considered simple enough to evaluate the effects that changes in the composition of each component could have on the studied bioactivities. The best linear regression models found, obtained through the Replacement Method variable subset selection method, are applied to predict 13 essential oils from a previous study with unknown property data. The results show that the simple methodology applied here could be useful for predicting properties of interest in complex mixtures such as essential oils.

POxload: Machine Learning Estimates Drug Loadings of Polymeric Micelles

Block copolymers, composed of poly(2-oxazoline)s and poly(2-oxazine)s, can serve as drug delivery systems; they form micelles that carry poorly water-soluble drugs. Many recent studies have investigated the effects of structural changes of the polymer and the hydrophobic cargo on drug loading. In this work, we combine these data to establish an extended formulation database. Different molecular properties and fingerprints are tested for their applicability to serve as formulation-specific mixture descriptors. A variety of classification and regression models are built for different descriptor subsets and thresholds of loading efficiency and loading capacity, with the best models achieving overall good statistics for both cross- and external validation (balanced accuracies of 0.8). Subsequently, important features are dissected for interpretation, and the DrugBank is screened for potential therapeutic use cases where these polymers could be used to develop novel formulations of hydrophobic drugs. The most promising models are provided as an open-source software tool for other researchers to test the applicability of these delivery systems for potential new drug candidates.

Mixtures Recomposition by Neural Nets: A Multidisciplinary Overview

Artificial Neural Networks (ANNs) are transforming how we understand chemical mixtures, providing an expressive view of the chemical space and multiscale processes. Their hybridization with physical knowledge can bridge the gap between predictivity and understanding of the underlying processes. This overview explores recent progress in ANNs, particularly their potential in the 'recomposition' of chemical mixtures. Graph-based representations reveal patterns among mixture components, and deep learning models excel in capturing complexity and symmetries when compared to traditional Quantitative Structure-Property Relationship models. Key components, such as Hamiltonian networks and convolution operations, play a central role in representing multiscale mixtures. The integration of ANNs with Chemical Reaction Networks and Physics-Informed Neural Networks for inverse chemical kinetic problems is also examined. The combination of sensors with ANNs shows promise in optical and biomimetic applications. A common ground is identified in the context of statistical physics, where ANN-based methods iteratively adapt their models by blending their initial states with training data. The concept of mixture recomposition unveils a reciprocal inspiration between ANNs and reactive mixtures, highlighting learning behaviors influenced by the training environment.

Integrated Computational and Experimental Framework for Inverse Screening of Candidate Antibacterial Nanomedicine

Nanomedicine is promising for disease prevention and treatment, but there are still many challenges that hinder its rapid development. A major challenge is to efficiently seek candidates with the desired therapeutic functions from tremendously available materials. Here, we report an integrated computational and experimental framework to seek alloy nanoparticles from the Materials Project library for antibacterial applications, aiming to learn the inverse screening concept from traditional medicine for nanomedicine. Because strong peroxidase-like catalytic activity and weak toxicity to normal cells are the desired material properties for antibacterial usage, computational screening implementing theoretical prediction models of catalytic activity and cytotoxicity is first conducted to select the candidates. Then, experimental screening based on scanning probe block copolymer lithography is used to verify and refine the computational screening results. Finally, the best candidate AuCu3 is synthesized in solution and its antibacterial performance over other nanoparticles against S. aureus and E. coli. is experimentally confirmed. The results show the power of inverse screening in accelerating the research and development of antibacterial nanomedicine, which may inspire similar strategies for other nanomedicines in the future.

Predicting the Performance of Functional Materials Composed of Polymeric Multicomponent Systems Using Artificial Intelligence-Formulations of Cleansing Foams as an Example

Cleansing foam is a common multicomponent polymeric functional material. It contains ingredients in innumerable combinations, which makes formulation optimization challenging. In this study, we used artificial intelligence (AI) with machine learning to develop a cleansing capability prediction system that considers the effects of self-assembled structures and chemical properties of ingredients. Over 500 cleansing foam samples were prepared and tested. Molecular descriptors and Hansen solubility index were used to estimate the cleansing capabilities of each formulation set. We used five machine-learning models to predict the cleansing capability. In addition, we employed an in silico formulation by generating virtual formulations and predicting their cleansing capabilities using an established AI model. The achieved accuracy was R2 = 0.770. Our observations revealed that mixtures of cosmetic ingredients exhibit complex interactions, resulting in nonlinear behavior, which adds to the complexity of predicting cleansing performance. Nevertheless, accurate chemical property descriptors, along with the aid of in silico formulations, enabled the identification of potential ingredients. We anticipate that our system will efficiently predict the chemical properties of polymer-containing blends.

Machine learning-based models for accessing thermal conductivity of liquids at different temperature conditions

Combating global warming-related climate change demands prompt actions to reduce greenhouse gas emissions, particularly carbon dioxide. Biomass-based biofuels represent a promising alternative fossil energy source. To convert biomass into energy, numerous conversion processes are performed at high pressure and temperature conditions, and the design and dimensioning of such processes requires thermophysical property data, particularly thermal conductivity, which are not always available in the literature. In this paper, we proposed the application of Chemoinformatics methodologies to investigate the prediction of thermal conductivity for hydrocarbons and oxygenated compounds. A compilation of experimental data followed by a careful data curation were performed to establish a database. The support vector machine algorithm has been applied to the database leading to models with good predictive abilities. The support vector regression (SVR) model has then been applied to an external set of compounds, i.e. not considered during the training of models. It showed that our SVR model can be used for the prediction of thermal conductivity values for temperatures and/or compounds that are not covered experimentally in the literature.

QSPR models to predict the physical hazards of mixtures: a state of art

Physical hazards of chemical mixtures, associated for example with their fire or explosion risks, are generally characterized using experimental tools. These tests can be expensive, complex, long to perform and even dangerous for operators. Therefore, for several years and especially with the implementation of the REACH regulation, predictive methods like quantitative structure-property relationships have been encouraged as alternatives tests to determine (eco)toxicological but also physical hazards of chemical substances. Initially, these approaches were intended for pure products, by considering a molecular similarity principle. However, additional to those for pure products, QSPR models for mixtures recently appeared and represent an increasing field of research. This study proposes a state of the art of existing QSPR models specifically dedicated to the prediction of the physical hazards of mixtures. Identified models have been analysed on the key elements of model development (experimental data and fields of application, descriptors used, development and validation methods). It draws up an overview of the potential and limitations of current models as well as areas of progress towards enlarged deployment as a complement to experimental characterizations, for example in the search for safer substances (according to safety-by-design concepts).

Converting Nanotoxicity Data to Information Using Artificial Intelligence and Simulation

Decades of nanotoxicology research have generated extensive and diverse data sets. However, data is not equal to information. The question is how to extract critical information buried in vast data streams. Here we show that artificial intelligence (AI) and molecular simulation play key roles in transforming nanotoxicity data into critical information, i.e., constructing the quantitative nanostructure (physicochemical properties)-toxicity relationships, and elucidating the toxicity-related molecular mechanisms. For AI and molecular simulation to realize their full impacts in this mission, several obstacles must be overcome. These include the paucity of high-quality nanomaterials (NMs) and standardized nanotoxicity data, the lack of model-friendly databases, the scarcity of specific and universal nanodescriptors, and the inability to simulate NMs at realistic spatial and temporal scales. This review provides a comprehensive and representative, but not exhaustive, summary of the current capability gaps and tools required to fill these formidable gaps. Specifically, we discuss the applications of AI and molecular simulation, which can address the large-scale data challenge for nanotoxicology research. The need for model-friendly nanotoxicity databases, powerful nanodescriptors, new modeling approaches, molecular mechanism analysis, and design of the next-generation NMs are also critically discussed. Finally, we provide a perspective on future trends and challenges.

Toxicity Assessment of the Binary Mixtures of Aquatic Organisms Based on Different Hypothetical Descriptors

Modern industrialization has led to the creation of a wide range of organic chemicals, especially in the form of multicomponent mixtures, thus making the evaluation of environmental pollution more difficult by normal methods. In this paper, we attempt to use forward stepwise multiple linear regression (MLR) and nonlinear radial basis function neural networks (RBFNN) to establish quantitative structure–activity relationship models (QSARs) to predict the toxicity of 79 binary mixtures of aquatic organisms using different hypothetical descriptors. To search for the proper mixture descriptors, 11 mixture rules were performed and tested based on preliminary modeling results. The statistical parameters of the best derived MLR model were Ntrain = 62, R2 = 0.727, RMS = 0.494, F = 159.537, Q2LOO = 0.727, and Q2pred = 0.725 for the training set; and Ntest = 17, R2 = 0.721, RMS = 0.508, F = 38.773, and q2ext = 0.720 for the external test set. The RBFNN model gave the following statistical results: Ntrain = 62, R2 = 0.956, RMS = 0.199, F = 1279.919, Q2LOO = 0.955, and Q2pred = 0.855 for the training set; and Ntest = 17, R2 = 0.880, RMS = 0.367, F = 110.980, and q2ext = 0.853 for the external test set. The quality of the models was assessed by validating the relevant parameters, and the final results showed that the developed models are predictive and can be used for the toxicity prediction of binary mixtures within their applicability domain.

Predicting the Surface Tension of Deep Eutectic Solvents: A Step Forward in the Use of Greener Solvents

Deep eutectic solvents (DES) are an important class of green solvents that have been developed as an alternative to toxic solvents. However, the large-scale industrial application of DESs requires fine-tuning their physicochemical properties. Among others, surface tension is one of such properties that have to be considered while designing novel DESs. In this work, we present the results of a detailed evaluation of Quantitative Structure-Property Relationships (QSPR) modeling efforts designed to predict the surface tension of DESs, following the Organization for Economic Co-operation and Development (OECD) guidelines. The data set used comprises a large number of structurally diverse binary DESs and the models were built systematically through rigorous validation methods, including ‘mixtures-out’- and ‘compounds-out’-based data splitting. The most predictive individual QSPR model found is shown to be statistically robust, besides providing valuable information about the structural and physicochemical features responsible for the surface tension of DESs. Furthermore, the intelligent consensus prediction strategy applied to multiple predictive models led to consensus models with similar statistical robustness to the individual QSPR model. The benefits of the present work stand out also from its reproducibility since it relies on fully specified computational procedures and on publicly available tools. Finally, our results not only guide the future design and screening of novel DESs with a desirable surface tension but also lays out strategies for efficiently setting up silico-based models for binary mixtures.

Mixtures-Inclusive In Silico Models of Ocular Toxicity Based on United States and International Hazard Categories

Computational modeling grounded in reliable experimental data can help design effective non-animal approaches to predict the eye irritation and corrosion potential of chemicals. The National Toxicology Program (NTP) Interagency Center for the Evaluation of Alternative Toxicological Methods (NICEATM) has compiled and curated a database of in vivo eye irritation studies from the scientific literature and from stakeholder-provided data. The database contains 810 annotated records of 593 unique substances, including mixtures, categorized according to UN GHS and US EPA hazard classifications. This study reports a set of in silico models to predict EPA and GHS hazard classifications for chemicals and mixtures, accounting for purity by setting thresholds of 100% and 10% concentration. We used two approaches to predict classification of mixtures: conventional and mixture-based. Conventional models evaluated substances based on the chemical structure of its major component. These models achieved balanced accuracy in the range of 68-80% and 87-96% for the 100% and 10% test concentration thresholds, respectively. Mixture-based models, which accounted for all known components in the substance by weighted feature averaging, showed similar or slightly higher accuracy of 72-79% and 89-94% for the respective thresholds. We also noted a strong trend between the pH feature metric calculated for each substance and its activity. Across all the models, the calculated pH of inactive substances was within one log10 unit of neutral pH, on average, while for active substances, pH varied from neutral by at least 2 log10 units. This pH dependency is especially important for complex mixtures. Additional evaluation on an external test set of 673 substances obtained from ECHA dossiers achieved balanced accuracies of 64-71%, which suggests that these models can be useful in screening compounds for ocular irritation potential. Negative predictive value was particularly high and indicates the potential application of these models in a bottom-up approach to identify nonirritant substances.

Deep Probabilistic Learning Model for Prediction of Ionic Liquids Toxicity

Identification of ionic liquids with low toxicity is paramount for applications in various domains. Traditional approaches used for determining the toxicity of ionic liquids are often expensive, and can be labor intensive and time consuming. In order to mitigate these limitations, researchers have resorted to using computational models. This work presents a probabilistic model built from deep kernel learning with the aim of predicting the toxicity of ionic liquids in the leukemia rat cell line (IPC-81). Only open source tools, namely, RDKit and Mol2vec, are required to generate predictors for this model; as such, its predictions are solely based on chemical structure of the ionic liquids and no manual extraction of features is needed. The model recorded an RMSE of 0.228 and R2 of 0.943. These results indicate that the model is both reliable and accurate. Furthermore, this model provides an accompanying uncertainty level for every prediction it makes. This is important because discrepancies in experimental measurements that generated the dataset used herein are inevitable, and ought to be modeled. A user-friendly web server was developed as well, enabling researchers and practitioners ti make predictions using this model.

New methods of top-to-down mixture toxicity prediction: A case study of eliminating of the effects of cosolvent from binary mixtures

At present, the toxicity prediction of mixtures mainly focuses on the concentration addition (CA) and independent action (IA) based on individual toxicants to predict the toxicity of multicomponent mixtures. This process of predicting the toxicity of multicomponent mixtures based on single substances or low component mixtures is called down-to-top method in this study. However, due to the particularity of some toxicants, we have to use the top-to-down idea to obtain or eliminate the toxicity of some components from mixtures. For example, the toxicity of toxicants is obtained from the toxicity of a mixture with, especially toxic, cosolvent added. In the study, two top-to-down methods, the inverse CA (ICA) and inverse IA (IIA) models, were proposed to eliminate the effects of a certain component from multicomponent mixtures. Furthermore, taking the eight binary mixtures consisting of different shapes of cosolvents (isopropyl alcohol (IPA) having hormesis and dimethyl sulfoxide (DMSO)) and toxicants (two ionic liquids and two pesticides) as an example, combined with the interaction evaluated by CA and IA model, the influence of different shapes of components on top-to-down toxicity prediction was explored. The results showed that cosolvent IPA having hormesis may cause unpredictable effects, even at low concentrations, and should be used with caution. For DMSO, most of the toxicant's toxicity obtained by ICA and IIA models were almost in accordance with those observed experimentally, which showed that ICA and IIA could effectively eliminate the effects of cosolvent, even if toxic cosolvent, from the mixture. Ultimately, a frame of cosolvent use and toxicity correction for the hydrophobic toxicant were suggested based on the top-to-down toxicity prediction method. The proposed methods improve the existing framework of mixture toxicity prediction and provide a new idea for mixture toxicity evaluation and risk assessment.

Chemical composition: Hearing insect defensive volatiles

Chemical signals mediate major ecological interactions in insects. However, using bioassays only, it is difficult to quantify the bioactivity of complex mixtures, such as volatile defensive secretions emitted by prey insects, and to assess the impact of single compounds on the repellence of the entire mixture. To represent chemical data in a different perceptive mode, we used a process of sonification by parameter mapping of single molecules, which translated chemical signals into acoustic signals. These sounds were then mixed at dB levels reflecting the relative concentrations of the molecules within species-specific secretions. Repellence of single volatiles, as well as mixtures of volatiles, against predators were significantly correlated with the repulsiveness of their respective auditory translates against humans, who mainly reacted to sound pressure. Furthermore, sound pressure and predator response were associated with the number of different molecules in a secretion. Our transmodal approach, from olfactory to auditory perception, offers further prospects for chemo-ecological research and data representation.

Selection of antileishmanial sesquiterpene lactones from SistematX database using a combined ligand-/structure-based virtual screening approach

Leishmaniasis refers to a complex of diseases, caused by the intracellular parasitic protozoans belonging to the genus Leishmania. Among the three types of disease manifestations, the most severe type is visceral leishmaniasis, which is caused by Leishmania donovani, and is diagnosed in more than 20,000 cases annually, worldwide. Because the current therapeutic options for disease treatment are associated with several limitations, the identification of new potential leads/drugs remains necessary. In this study, a combined approach was used, based on two different virtual screening (VS) methods, which were designed to select promising antileishmanial agents from among the entire sesquiterpene lactone (SL) dataset registered in SistematX, a web interface for managing a secondary metabolite database that is accessible by multiple platforms on the Internet. Thus, a ChEMBL dataset, including 3159 and 1569 structures that were previously tested against L. donovani amastigotes and promastigotes in vitro, respectively, was used to develop two random forest models, which performed with greater than 74% accuracy in both the cross-validation and test sets. Subsequently, a ligand-based VS assay was performed against the 1306 SistematX-registered SLs. In parallel, the crystal structures of three L. donovani target proteins, N-myristoyltransferase, ornithine decarboxylase, and mitogen-activated protein kinase 3, and a homology model of pteridine reductase 1 were used to perform a structure-based VS, using molecular docking, of the entire SistematX SL dataset. The consensus analysis of these two VS approaches resulted in the normalization of probability scores and identified 13 promising, enzyme-targeting, antileishmanial SLs from SistematX that may act against L. donovani. A combined approach based on two different virtual screening methods (structure-based and ligand-based) was performed using an in-house dataset composed of 1306 sesquiterpene lactones to identify potential antileishmanial (Leishmania donovani) structures.

Density of Deep Eutectic Solvents: The Path Forward Cheminformatics-Driven Reliable Predictions for Mixtures

Deep eutectic solvents (DES) are often regarded as greener sustainable alternative solvents and are currently employed in many industrial applications on a large scale. Bearing in mind the industrial importance of DES—and because the vast majority of DES has yet to be synthesized—the development of cheminformatic models and tools efficiently profiling their density becomes essential. In this work, after rigorous validation, quantitative structure-property relationship (QSPR) models were proposed for use in estimating the density of a wide variety of DES. These models were based on a modelling dataset previously employed for constructing thermodynamic models for the same endpoint. The best QSPR models were robust and sound, performing well on an external validation set (set up with recently reported experimental density data of DES). Furthermore, the results revealed structural features that could play crucial roles in ruling DES density. Then, intelligent consensus prediction was employed to develop a consensus model with improved predictive accuracy. All models were derived using publicly available tools to facilitate easy reproducibility of the proposed methodology. Future work may involve setting up reliable, interpretable cheminformatic models for other thermodynamic properties of DES and guiding the design of these solvents for applications.

Simplex representation of molecular structure as universal QSAR/QSPR tool

We review the development and application of the Simplex approach for the solution of various QSAR/QSPR problems. The general concept of the simplex method and its varieties are described. The advantages of utilizing this methodology, especially for the interpretation of QSAR/QSPR models, are presented in comparison to other fragmentary methods of molecular structure representation. The utility of SiRMS is demonstrated not only in the standard QSAR/QSPR applications, but also for mixtures, polymers, materials, and other complex systems. In addition to many different types of biological activity (antiviral, antimicrobial, antitumor, psychotropic, analgesic, etc.), toxicity and bioavailability, the review examines the simulation of important properties, such as water solubility, lipophilicity, as well as luminescence, and thermodynamic properties (melting and boiling temperatures, critical parameters, etc.). This review focuses on the stereochemical description of molecules within the simplex approach and details the possibilities of universal molecular stereo-analysis and stereochemical configuration description, along with stereo-isomerization mechanism and molecular fragment "topography" identification.

Development of an Infinite Dilution Activity Coefficient Prediction Model for Organic Solutes in Ionic Liquids with Modified Partial Equalization Orbital Electronegativity Method Derived Descriptors

The objective of this study was to develop a robust prediction model for the infinite dilution activity coefficients (γ ∞) of organic molecules in diverse ionic liquid (IL) solvents. Electrostatic, hydrogen bond, polarizability, molecular structure, and temperature terms were used in model development. A feed-forward model based on artificial neural networks was developed with 34,754 experimental activity coefficients, a combination of 195 IL solvents (88 cations and 38 anions), and 147 organic solutes at a temperature range of 298 to 408 K. The root mean squared error (RMSE) of the training set and test set was 0.219 and 0.235, respectively. The R 2 of the training set and the test set was 0.984 and 0.981, respectively. The applicability domain was determined through a Williams plot, which implied that water and halogenated compounds were outside of the applicability domain. The robustness test shows that the developed model is robust. The web server supports using the developed prediction model and is freely available at https://preadmet.bmdrc.kr/activitycoefficient_mainpage/prediction/.

Identification of Kaurane-Type Diterpenes as Inhibitors of Leishmania Pteridine Reductase I

The current treatments against Leishmania parasites present high toxicity and multiple side effects, which makes the control and elimination of leishmaniasis challenging. Natural products constitute an interesting and diverse chemical space for the identification of new antileishmanial drugs. To identify new drug options, an in-house database of 360 kauranes (tetracyclic diterpenes) was generated, and a combined ligand- and structure-based virtual screening (VS) approach was performed to select potential inhibitors of Leishmania major (Lm) pteridine reductase I (PTR1). The best-ranked kauranes were employed to verify the validity of the VS approach through LmPTR1 enzyme inhibition assay. The half-maximal inhibitory concentration (IC₅₀) values of selected bioactive compounds were examined using the random forest (RF) model (i.e., 2β-hydroxy-menth-6-en-5β-yl ent-kaurenoate (135) and 3α-cinnamoyloxy-ent-kaur-16-en-19-oic acid (302)) were below 10 μM. A compound similar to 302, 3α-p-coumaroyloxy-ent-kaur-16-en-19-oic acid (302a), was also synthesized and showed the highest activity against LmPTR1. Finally, molecular docking calculations and molecular dynamics simulations were performed for the VS-selected, most-active kauranes within the active sites of PTR1 hybrid models, generated from three Leishmania species that are known to cause cutaneous leishmaniasis in the new world (i.e., L. braziliensis, L. panamensis, and L. amazonensis) to explore the targeting potential of these kauranes to other species-dependent variants of this enzyme.

Functional and Material Properties in Nanocatalyst Design: A Data Handling and Sharing Problem

(1) Background: Properties and descriptors are two forms of molecular in silico representations. Properties can be further divided into functional, e.g., catalyst or drug activity, and material, e.g., X-ray crystal data. Millions of real measured functional property records are available for drugs or drug candidates in online databases. In contrast, there is not a single database that registers a real conversion, TON or TOF data for catalysts. All of the data are molecular descriptors or material properties, which are mainly of a calculation origin. (2) Results: Here, we explain the reason for this. We reviewed the data handling and sharing problems in the design and discovery of catalyst candidates particularly, material informatics and catalyst design, structural coding, data collection and validation, infrastructure for catalyst design and the online databases for catalyst design. (3) Conclusions: Material design requires a property prediction step. This can only be achieved based on the registered real property measurement. In reality, in catalyst design and discovery, we can observe either a severe functional property deficit or even property famine.

QSAR models for insecticidal properties of plant essential oils on the housefly (Musca domestica L.)

The fumigant and topical activities exhibited by 27 plant-derived essentials oils (EOs) on adult M. domestica housefly are predicted through the Quantitative Structure-Activity Relationship (QSAR) theory. These molecular structure based calculations are performed on 253 structurally diverse compounds from the EOs, where the number of constituents in each essential oil mixture varies between 2 to 24. A large number of 86,048 non-conformational mixture descriptors are derived as linear combinations of the molecular descriptors of the EO components. Two strategies are compared for the mixture descriptor formulation, which consider or avoid the use of the chemical composition. The multivariable linear regression QSAR models of the present work are useful for fumigant and topical applications, describing predictive parallelisms for the insecticidal activity of the analysed complex mixtures.

Curated Data In - Trustworthy In Silico Models Out: The Impact of Data Quality on the Reliability of Artificial Intelligence Models as Alternatives to Animal Testing

New Approach Methodologies (NAMs) that employ artificial intelligence (AI) for predicting adverse effects of chemicals have generated optimistic expectations as alternatives to animal testing. However, the major underappreciated challenge in developing robust and predictive AI models is the impact of the quality of the input data on the model accuracy. Indeed, poor data reproducibility and quality have been frequently cited as factors contributing to the crisis in biomedical research, as well as similar shortcomings in the fields of toxicology and chemistry. In this article, we review the most recent efforts to improve confidence in the robustness of toxicological data and investigate the impact that data curation has on the confidence in model predictions. We also present two case studies demonstrating the effect of data curation on the performance of AI models for predicting skin sensitisation and skin irritation. We show that, whereas models generated with uncurated data had a 7-24% higher correct classification rate (CCR), the perceived performance was, in fact, inflated owing to the high number of duplicates in the training set. We assert that data curation is a critical step in building computational models, to help ensure that reliable predictions of chemical toxicity are achieved through use of the models.

Toxicity of binary mixtures of pesticides and pharmaceuticals toward Vibrio fischeri: Assessment by quantitative structure-activity relationships

Pollutants in real aquatic systems commonly occur as chemical mixtures. Yet, the corresponding risk assessment is still mostly based on information on single-pollutant toxicity, accepting the assumption that pollutant mixtures exhibit additive toxicity effect which is often not the case. Therefore, it is still better to use the experimental approach. Unfortunately, experimental determination of toxicity for each mixture is practically unfeasible. In this study, quantitative structure-activity relationship (QSAR) models for the prediction of toxicity of binary mixtures towards bioluminescent bacteria Vibrio fischeri were developed at three toxicity levels (EC₁₀, EC₃₀ and EC₅₀). For model development, experimentally determined toxicity values of 14 pollutants (pharmaceuticals and pesticides) were correlated with their structural features, applying multiple linear regression together with genetic algorithm. Statistical analysis, internal validation and external validation of the models were carried out. The toxicity is accurately predicted by all three models. EC₃₀ and EC₅₀ values are mostly influenced by geometrical distances between nitrogen and sulfur atoms. Furthermore, the simultaneous presence of oxygen and chlorine atoms in mixture can induce the increase in toxicity. At lower effect levels (EC₁₀), nitrogen atom bonded to different groups has the highest impact on mixture toxicity. Thus, the analysis of the descriptors involved in the developed models can give insight into toxic mechanisms of the binary systems.

Synergistic and Antagonistic Drug Combinations against SARS-CoV-2

Antiviral drug development for coronavirus disease 2019 (COVID-19) is occurring at an unprecedented pace, yet there are still limited therapeutic options for treating this disease. We hypothesized that combining drugs with independent mechanisms of action could result in synergy against SARS-CoV-2, thus generating better antiviral efficacy. Using in silico approaches, we prioritized 73 combinations of 32 drugs with potential activity against SARS-CoV-2 and then tested them in vitro. Sixteen synergistic and eight antagonistic combinations were identified; among 16 synergistic cases, combinations of the US Food and Drug Administration (FDA)-approved drug nitazoxanide with remdesivir, amodiaquine, or umifenovir were most notable, all exhibiting significant synergy against SARS-CoV-2 in a cell model. However, the combination of remdesivir and lysosomotropic drugs, such as hydroxychloroquine, demonstrated strong antagonism. Overall, these results highlight the utility of drug repurposing and preclinical testing of drug combinations for discovering potential therapies to treat COVID-19.

Chemoinformatics for the Safety of Energetic and Reactive Materials at Ineris

The characterization of physical hazards of substances is a key information to manage the risks associated to their use, storage and transport. With decades of work in this area, Ineris develops and implements cutting-edge experimental facilities allowing such characterizations at different scales and under various conditions to study all of the dreaded accident scenarios. This review presents the efforts engaged by Ineris more recently in the field of chemoinformatics to develop and use new predictive methods for the anticipation and management of industrials risks associated to energetic and reactive materials as a complement to experiments. An overview of the methods used for the development of Quantitative Structure-Property Relationships for physical hazards are presented and discussed regarding the specificities associated to this class of properties. A review of models developed at Ineris is also provided from the first tentative models on the explosivity of nitro compounds to the successful application to the flammability of organic mixtures. Then, a discussion is proposed on the use of QSPR models. Good practices for robust use for QSPR models are recalled with specific comments related to physical hazards, notably for regulatory purpose. Dissemination and training efforts engaged by Ineris are also presented. The potential offered by these predictive methods in terms of in silico design and for the development of new intrinsically safer technologies in safety-by-design strategies is finally discussed. At last, challenges and perspectives to extend the application of chemoinformatics in the field of safety and in particular for the physical hazards of energetic and reactive substances are proposed.

BNNmix: A new approach for predicting the mixture toxicity of multiple components based on the back-propagation neural network

The chemical mixtures in various environmental media not only have concentration diversity but also mixture-ratio diversity. It is impossible to experimentally determine the toxicities of all mixtures; therefore, it is necessary to develop effective methods based on models to predict mixture toxicity. In this study, a new approach (BNNmix) based on the back-propagation neural network (BPNN) was developed and used to predict the toxicities of seven-component mixtures (consisting of two substituted phenols, two pesticides, two ionic liquids, and one heavy metal) on Caenorhabditis elegans. We found that the combined toxicities of various mixtures used in the experiments were neither global concentration-additive nor global response-additive, which implied that it was impossible to accurately predict the toxicities of such mixtures by using common models such as concentration addition (CA) and response addition (independent action, IA). Using the BNNmix approach to estimate or predict the toxicities of the mixtures under test, it was found that the predictive toxicities of various mixtures with different mixture ratios and concentrations were almost in accordance with those observed experimentally. Unlike the CA and IA models, the BNNmix approach can predict not only the toxicities of mixtures having toxicological interactions but also those with global concentration or response additivities.

Discovery of Synergistic and Antagonistic Drug Combinations against SARS-CoV-2 In Vitro

COVID-19 is undoubtedly the most impactful viral disease of the current century, afflicting millions worldwide. As yet, there is not an approved vaccine, as well as limited options from existing drugs for treating this disease. We hypothesized that combining drugs with independent mechanisms of action could result in synergy against SARS-CoV-2. Using in silico approaches, we prioritized 73 combinations of 32 drugs with potential activity against SARS-CoV-2 and then tested them in vitro . Overall, we identified 16 synergistic and 8 antagonistic combinations, 4 of which were both synergistic and antagonistic in a dose-dependent manner. Among the 16 synergistic cases, combinations of nitazoxanide with three other compounds (remdesivir, amodiaquine and umifenovir) were the most notable, all exhibiting significant synergy against SARS-CoV-2. The combination of nitazoxanide, an FDA-approved drug, and remdesivir, FDA emergency use authorization for the treatment of COVID-19, demonstrate a strong synergistic interaction. Notably, the combination of remdesivir and hydroxychloroquine demonstrated strong antagonism. Overall, our results emphasize the importance of both drug repurposing and preclinical testing of drug combinations for potential therapeutic use against SARS-CoV-2 infections.

QSAR without borders

Prediction of chemical bioactivity and physical properties has been one of the most important applications of statistical and more recently, machine learning and artificial intelligence methods in chemical sciences. This field of research, broadly known as quantitative structure-activity relationships (QSAR) modeling, has developed many important algorithms and has found a broad range of applications in physical organic and medicinal chemistry in the past 55+ years. This Perspective summarizes recent technological advances in QSAR modeling but it also highlights the applicability of algorithms, modeling methods, and validation practices developed in QSAR to a wide range of research areas outside of traditional QSAR boundaries including synthesis planning, nanotechnology, materials science, biomaterials, and clinical informatics. As modern research methods generate rapidly increasing amounts of data, the knowledge of robust data-driven modelling methods professed within the QSAR field can become essential for scientists working both within and outside of chemical research. We hope that this contribution highlighting the generalizable components of QSAR modeling will serve to address this challenge.

Representation of the Structure-A Key Point of Building QSAR/QSPR Models for Ionic Liquids

The process of encoding the structure of chemicals by molecular descriptors is a crucial step in quantitative structure-activity/property relationships (QSAR/QSPR) modeling. Since ionic liquids (ILs) are disconnected structures, various ways of representing their structure are used in the QSAR studies: the models can be based on descriptors either derived for particular ions or for the whole ionic pair. We have examined the influence of the type of IL representation (separate ions vs. ionic pairs) on the model's quality, the process of the automated descriptors selection and reliability of the applicability domain (AD) assessment. The result of the benchmark study showed that a less precise description of ionic liquid, based on the 2D descriptors calculated for ionic pairs, is sufficient to develop a reliable QSAR/QSPR model with the highest accuracy in terms of calibration as well as validation. Moreover, the process of a descriptors' selection is more effective when the possible number of variables can be decreased at the beginning of model development. Additionally, 2D descriptors usually demand less effort in mechanistic interpretation and are more convenient for virtual screening studies.

Predicting acute contact toxicity of organic binary mixtures in honey bees (A. mellifera) through innovative QSAR models

Pollinators such as honey bees are of considerable importance, because of the crucial pollination services they provide for food crops and wild plants. Since bees are exposed to a wide range of multiple chemicals "mixtures" both of anthropogenic (e.g. plant protection products) and natural origin (e.g. plant toxins), understanding their combined toxicity is critical. Although honey bees are employed worldwide as surrogate species for Apis and non-Apis bees in toxicity tests, it is practically unfeasible to perform in vivo tests for all mixtures of chemicals. Therefore, Quantitative Structure-Activity Relationships (QSAR) models can be developed using available data and can provide useful tools to predict such combined toxicity. Here, three different QSAR models within the CORAL software have been calibrated and validated for honey bees (A. mellifera) to predict the acute contact mixtures potency (LD_50-mix), in two regression based-models, and the nature of combined toxicity (synergism / non-synergism) in a classification-based model. Experimental data on binary mixtures (n = 123) (LD_50-mix) including dose response data (n = 97) and corresponding Toxic Unit values were retrieved from EFSA databases. The models were built using the principle of extraction of attributes from SMILES (or quasi-SMILES) while calculating so-called correlation weights for these attributes using Monte Carlo techniques. The two regression models were validated for their reliability and robustness (R² = 0.89, CCC = 0.92, Q² = 0.81; R² = 0.87, CCC = 0.89, Q² = 0.75). The classification model was validated using sensitivity (=0.86), specificity (=1), accuracy (=0.96), and Matthews correlation coefficient (MCC = 0.90) as qualitative statistical validation parameters. Results indicate that these QSAR models successfully predict acute contact toxicity of binary mixtures in honey bees and can support prioritisation of multiple chemicals of concerns. Data gaps and further development of QSAR models for honey bees are highlighted particularly for chronic and sub-lethal effects.

Assessment of the cardiovascular adverse effects of drug-drug interactions through a combined analysis of spontaneous reports and predicted drug-target interactions

Adverse drug effects (ADEs) are one of the leading causes of death in developed countries and are the main reason for drug recalls from the market, whereas the ADEs that are associated with action on the cardiovascular system are the most dangerous and widespread. The treatment of human diseases often requires the intake of several drugs, which can lead to undesirable drug-drug interactions (DDIs), thus causing an increase in the frequency and severity of ADEs. An evaluation of DDI-induced ADEs is a nontrivial task and requires numerous experimental and clinical studies. Therefore, we developed a computational approach to assess the cardiovascular ADEs of DDIs. This approach is based on the combined analysis of spontaneous reports (SRs) and predicted drug-target interactions to estimate the five cardiovascular ADEs that are induced by DDIs, namely, myocardial infarction, ischemic stroke, ventricular tachycardia, cardiac failure, and arterial hypertension. We applied a method based on least absolute shrinkage and selection operator (LASSO) logistic regression to SRs for the identification of interacting pairs of drugs causing corresponding ADEs, as well as noninteracting pairs of drugs. As a result, five datasets containing, on average, 3100 potentially ADE-causing and non-ADE-causing drug pairs were created. The obtained data, along with information on the interaction of drugs with 1553 human targets predicted by PASS Targets software, were used to create five classification models using the Random Forest method. The average area under the ROC curve of the obtained models, sensitivity, specificity and balanced accuracy were 0.837, 0.764, 0.754 and 0.759, respectively. The predicted drug targets were also used to hypothesize the potential mechanisms of DDI-induced ventricular tachycardia for the top-scoring drug pairs. The created five classification models can be used for the identification of drug combinations that are potentially the most or least dangerous for the cardiovascular system.

Predicting Synergism of Cancer Drug Combinations Using NCI-ALMANAC Data

Drug combinations are of great interest for cancer treatment. Unfortunately, the discovery of synergistic combinations by purely experimental means is only feasible on small sets of drugs. In silico modeling methods can substantially widen this search by providing tools able to predict which of all possible combinations in a large compound library are synergistic. Here we investigate to which extent drug combination synergy can be predicted by exploiting the largest available dataset to date (NCI-ALMANAC, with over 290,000 synergy determinations). Each cell line is modeled using primarily two machine learning techniques, Random Forest (RF) and Extreme Gradient Boosting (XGBoost), on the datasets provided by NCI-ALMANAC. This large-scale predictive modeling study comprises more than 5,000 pair-wise drug combinations, 60 cell lines, 4 types of models, and 5 types of chemical features. The application of a powerful, yet uncommonly used, RF-specific technique for reliability prediction is also investigated. The evaluation of these models shows that it is possible to predict the synergy of unseen drug combinations with high accuracy (Pearson correlations between 0.43 and 0.86 depending on the considered cell line, with XGBoost providing slightly better predictions than RF). We have also found that restricting to the most reliable synergy predictions results in at least 2-fold error decrease with respect to employing the best learning algorithm without any reliability estimation. Alkylating agents, tyrosine kinase inhibitors and topoisomerase inhibitors are the drugs whose synergy with other partner drugs are better predicted by the models. Despite its leading size, NCI-ALMANAC comprises an extremely small part of all conceivable combinations. Given their accuracy and reliability estimation, the developed models should drastically reduce the number of required in vitro tests by predicting in silico which of the considered combinations are likely to be synergistic.

A chemoinformatics approach for the characterization of hybrid nanomaterials: safer and efficient design perspective

In this study, photocatalytic properties and in vitro cytotoxicity of 29 TiO2-based multi-component nanomaterials (i.e., hybrids of more than two composition types of nanoparticles) were evaluated using a combination of the experimental testing and supervised machine learning modeling. TiO2-based multi-component nanomaterials with metal clusters of silver, and their mixtures with gold, palladium, and platinum were successfully synthesized. Two activities, photocatalytic activity and cytotoxicity, were studied. A novel cheminformatic approach was developed and applied for the computational representation of the photocatalytic activity and cytotoxicity effect. In this approach, features of investigated TiO2-based hybrid nanomaterials were reflected by a series of novel additive descriptors for hybrid and hybrid nanostructures (denoted as "hybrid nanosctructure descriptors"). These descriptors are based on quantum chemical calculations and the Smoluchowski equation. The obtained experimental data and calculated hybrid-nanostructure descriptors were used to develop novel predictive Quantitative Structure-Activity Relationship computational models (called "nano-QSARmix"). The proposed modeling approach is an initial step in the understanding of the relationships between physicochemical properties of hybrid nanoparticles, their toxicity, and photochemical activity under UV-vis irradiation. Acquired knowledge supports the safe-by-design approaches relevant to the development of efficient hybrid nanomaterials with reduced hazardous effects.

Cheminformatics-driven discovery of polymeric micelle formulations for poorly soluble drugs

Many drug candidates fail therapeutic development because of poor aqueous solubility. We have conceived a computer-aided strategy to enable polymeric micelle-based delivery of poorly soluble drugs. We built models predicting both drug loading efficiency (LE) and loading capacity (LC) using novel descriptors of drug-polymer complexes. These models were employed for virtual screening of drug libraries, and eight drugs predicted to have either high LE and high LC or low LE and low LC were selected. Three putative positives, as well as three putative negative hits, were confirmed experimentally (implying 75% prediction accuracy). Fortuitously, simvastatin, a putative negative hit, was found to have the desired micelle solubility. Podophyllotoxin and simvastatin (LE of 95% and 87% and LC of 43% and 41%, respectively) were among the top five polymeric micelle-soluble compounds ever studied experimentally. The success of the strategy described herein suggests its broad utility for designing drug delivery systems.

Prediction of the Auto-Ignition Temperatures of Binary Miscible Liquid Mixtures from Molecular Structures

A quantitative structure-property relationship (QSPR) study is performed to predict the auto-ignition temperatures (AITs) of binary liquid mixtures based on their molecular structures. The Simplex Representation of Molecular Structure (SiRMS) methodology was employed to describe the structure characteristics of a series of 132 binary miscible liquid mixtures. The most rigorous “compounds out” strategy was employed to divide the dataset into the training set and test set. The genetic algorithm (GA) combined with multiple linear regression (MLR) was used to select the best subset of SiRMS descriptors, which significantly contributes to the AITs of binary liquid mixtures. The result is a multilinear model with six parameters. Various strategies were employed to validate the developed model, and the results showed that the model has satisfactory robustness and predictivity. Furthermore, the applicability domain (AD) of the model was defined. The developed model could be considered as a new way to reliably predict the AITs of existing or new binary miscible liquid mixtures, belonging to its AD.

Modeling of Glass Transition Temperatures for Polymeric Coating Materials: Application of QSPR Mixture-based Approach

Cross-linked block copolymers are structurally complex, and utilization of traditional methods of molecular representation in chemoinformatics is only of limited applicability. Therefore, we introduced new techniques of structural representation for block copolymers. We developed additive and combinatorial approaches that treat a copolymer as a mixture system. In this approach, DRAGON descriptors are concentration-weighted for all chemicals in the reaction mixture. As a proof of concept, we have studied glass transition temperatures of block copolymers of hydroxyalkyl- and dihydroxyalkyl carbamate terminated poly(dimethylsiloxane) oligomers with poly(-caprolactone) and developed four quantitative structure-property relationships (QSPR) models. The correlation coefficient (R² ) for mentioned QSPR models ranges from 0.851 to 0.911 for the training set. In addition to the newly introduced technique we found that the octanol-water partition coefficient and 3D-MoRSE unweighted descriptors were the most important descriptors for the studied property. The results of the study demonstrated that all chemicals in reaction mixture influenced the glass transition temperatures.

Prediction of Lower Flammability Limits for Binary Hydrocarbon Gases by Quantitative Structure-A Property Relationship Approach

The lower flammability limit (LFL) is one of the most important parameters for evaluating the fire and explosion hazards of flammable gases or vapors. This study proposed quantitative structure-property relationship (QSPR) models to predict the LFL of binary hydrocarbon gases from their molecular structures. Twelve different mixing rules were employed to derive mixture descriptors for describing the structures characteristics of a series of 181 binary hydrocarbon mixtures. Genetic algorithm (GA)-based multiple linear regression (MLR) was used to select the most statistically effective mixture descriptors on the LFL of binary hydrocarbon gases. A total of 12 multilinear models were obtained based on the different mathematical formulas. The best model, issued from the norm of the molar contribution formula, was achieved as a six-parameter model. The best model was then rigorously validated using multiple strategies and further extensively compared to the previously published model. The results demonstrated the robustness, validity, and satisfactory predictivity of the proposed model. The applicability domain (AD) of the model was defined as well. The proposed best model would be expected to present an alternative to predict the LFL values of existing or new binary hydrocarbon gases, and provide some guidance for prioritizing the design of safer blended gases with desired properties.

Universal nanohydrophobicity predictions using virtual nanoparticle library

To facilitate the development of new nanomaterials, especially nanomedicines, a novel computational approach was developed to precisely predict the hydrophobicity of gold nanoparticles (GNPs). The core of this study was to develop a large virtual gold nanoparticle (vGNP) library with computational nanostructure simulations. Based on the vGNP library, a nanohydrophobicity model was developed and then validated against externally synthesized and tested GNPs. This approach and resulted model is an efficient and effective universal tool to visualize and predict critical physicochemical properties of new nanomaterials before synthesis, guiding nanomaterial design.

QSAR Study of Biologically Active Essential Oils against Beetles Infesting the Walnut in Catamarca, Argentina

Essential oils from six species of aromatic plants collected in the Catamarca Province of Argentina were evaluated for their chemical composition and repellent and insecticidal activities against beetles of the genus Carpophilus (Coleoptera: Nitidulidae) and Oryzaephilus (Coleoptera: Silvanidae) that infest the local walnut production. Experimental data were analyzed using generalized estimating equations, with normal distribution and the identity link function. From the spectral information from the tested essential oils, we worked their molecular modeling as mixtures by developing mixture descriptors ( D_mix) that combined the molecular descriptor of each component in the mixture ( d _i) and its relative concentration ( x _i), i.e., D_mix = f( d _i, x _i). The application of chemoinformatic approaches determined that a combination of mixture descriptors related to molecular size, branchedness, charge distribution, and electronegativity were useful to explain the bioactivity profile against Carpophilus spp. and Oryzaephilus spp. The reported models were rigorously validated using stringent statistical parameters and essential oils reported with repellent activity against other beetle species from the Nitidulidae and Silvanidae families. This model confirmed each essential oil as a repellent with a comparable performance to the experimental reports.

Best Practices for QSAR Model Reporting: Physical and Chemical Properties, Ecotoxicity, Environmental Fate, Human Health, and Toxicokinetics Endpoints

BACKGROUND

Quantitative and qualitative structure–activity relationships (QSARs) have been used to understand chemical behavior for almost a century. The main source of QSAR models is the scientific literature, but the open question is how well these models are documented.

OBJECTIVES

The main aim of this study was to critically analyze the publication practices of QSARs with regard to transparency, potential reproducibility, and independent verification. The focus was on the level of technical completeness of the published QSARs.

METHODS

A total of 1,533 QSAR articles reporting 79 individual endpoints, mostly in environmental and health science, were reviewed. The QSAR parameters required for technical completeness were grouped into five categories: chemical structures, experimental endpoint values, descriptor values, mathematical representation of the model, and predicted endpoint values. The data were summarized and discussed using Circos plots.

RESULTS

Altogether, 42.5% of the reviewed articles were found to be potentially reproducible. The potential reproducibility for different endpoint groups varied; the respective rates were 39% for physical and chemical properties, 52% for ecotoxicity, 56% for environmental fate, 30% for human health, and 32% for toxicokinetics. The reproducibility of QSARs is discussed and placed in the context of the reproducibility of the experimental methods. Included are 65 references to open QSAR datasets as examples of models restored from scientific articles.

DISCUSSION

Strikingly poor documentation of QSARs was observed, which reduces the transparency, availability, and consequently, the application of research results in scientific, industrial, and regulatory areas. A list of the components needed to ensure the best practices for QSAR reporting is provided, allowing long-term use and preservation of the models. This list also allows an assessment of the reproducibility of models by interested parties such as journal editors, reviewers, regulators, evaluators, and potential users. https://doi.org/10.1289/EHP3264.