QSAR Research Articles

The role of physicochemical and topological parameters in drug design

Quantitative structure activity relationship (QSAR) is a widely used tool in rational drug design that establishes relationships between the physicochemical and topological descriptors of ligands and their biological activities. Obtained QSAR models help identify descriptors that play pivotal roles in the biological activity of ligands. This not only helps the prediction of new compounds with desirable biological activities but also helps with the design of new compounds with better activities and low toxicities. QSAR commonly uses lipophilicity (logP), hydrophobicity (logD), water solubility (logS), the acid–base dissociation constant (pKa), the dipole moment, the highest occupied molecular orbital (HOMO), the lowest unoccupied molecular orbital (LUMO), molecular weight (MW), molar volume (MV), molar refractivity (MR), and the kappa index as physicochemical parameters. Some commonly used topological indices in QSAR are the Wiener index, Platt index, Hosoya index, Zagreb indices, Balaban index, and E-state index. This review presents a brief description of the significance of the most extensively used physicochemical and topological parameters in drug design.

Efficient photocatalytic degradation of ethylenethiourea with zinc oxide photocatalysts: Parameters optimization, mechanism insight, exploration of degradation pathways and toxicity assessment

Traditional advanced oxidation processes (AOPs) for the removal of toxic and refractory ethylenethiourea (ETU) is uneconomical because of the demand for continuous chemical input and full-scale systems. Herein, ETU was efficiently mineralized without sacrificial reagents for the first time through the utilization of ZnO nano-photocatalysts which were synthesized by the adjustment of ratio between the precursor ZnCl2 and alkaline source KOH during hydrothermal process. Compared to other ZnO samples, ZnO-1/3 (1:3 molar (M) ratio of ZnCl2 to KOH) exhibited higher apparent rate constant (0.00661 min−1) towards the degradation of ETU because of more effective electron-hole separation. The optimum conditions for ETU degradation (ZnO-1/3 dosage of 0.4 g/L, initial ETU concentration of 10 mg/L and photocatalytic reaction temperature of 35 °C) were acquired according to the outcome of the response surface methodology (RSM) based on the Box-Behnken design (BBD), and the predicted maximum removal efficiency of ETU reached 93.17 % and was consistent with the experimental results under this optimum conditions. The h+ and •OH were proved to the main reactive species during the photocatalytic process by the free radical capture experiments and electron spin resonance (ESR) analysis. Furthermore, the possible degradation pathways of ETU were suggested based on the identified degradation intermediates, and the reduced toxicity of intermediates was also demonstrated through the quantitative structure–activity relationship (QSAR) prediction and the growth of mung bean seedlings. Hence, photocatalysis with ZnO can be regarded as a promising alternative for ETU elimination.

Coordination Number Regulation of Iron Single‐Atom Nanozyme for Enhancing H2O2 Activation and Selectively Eliminating Cephalosporin Antibiotics

AbstractNanozymes present promising alternatives to natural enzymes, but controlling nanozymes' performance and employing them for selectively removing antibiotics are extremely challenging. Employing theoretical calculations to design the coordination environments of mental and coordination atoms for directing single‐atom nanozymes synthesis emerges as a promising strategy to enhance their efficiency and selectivity in antibiotic elimination. In this study, the peroxidase‐like specificity of iron single‐atom nanozymes (FeSA‐Nx, x = 2,3, and 4) with specific Fe–N coordination numbers is demonstrated based on theoretical calculations. These calculations guide the synthesis of corresponding ultra‐thin FeSA‐Nx, achieving a high degree of consistency between theoretical predictions and experimental results. FeSA‐N3 with a Fe─N3 coordination number proves to be the most effective. The efficient electron transfer from Fe─N3 site to H2O2 reduces the free energy required for •OH generation. Quantitative structure‐activity relationship (QSAR) analysis reveals a strong positive correlation between degradation efficiency of cephalosporins and their electron‐donating capabilities (R2 = 0.820–0.929), realizing selectively eliminating cephalosporins. Integration FeSA‐N3 into ceramic membrane (FeSA‐N3/CM) improves hydrophilicity, achieving continuous and stable removal of cephalosporin. This study provides valuable insights into coordination number regulating nanozyme properties for selective antibiotics removal and offers novel perspectives for FeSA‐N3 application in integrated systems.

Prediction and Interpretability Study of the Glass Transition Temperature of Polyimide Based on Machine Learning with Quantitative Structure-Property Relationship (Tg-QSPR).

The glass transition temperature (Tg) is a crucial characteristic of polyimides (PIs). Developing a Tg predictive model using machine learning methodologies can facilitate the design of PI structures and expedite the development process. In this investigation, a data set comprising 1257 PIs was assembled, with Tg values determined using differential scanning calorimetry. 210 molecular descriptors were computed using RDKit, and subsequently, six distinct feature selection methodologies were employed to refine the descriptor set. Quantitative structure-property relationship models targeting Tg (Tg-QSPR) were then constructed using five ensemble learning algorithms and one deep learning algorithm. These models exhibited high predictive accuracy and robustness, with the CATBoost model demonstrating the highest accuracy, achieving a coefficient of determination of 0.823 for the test set, a mean absolute error of 20.1 °C, and a root-mean-square error of 29.0 °C. The study identified the NumRotatableBonds descriptor as particularly influential on Tg, showing a negative correlation with the property. Additionally, the model's accuracy was validated using ten new PI films not included in the original data set, resulting in absolute errors ranging from 2.5 to 26.9 °C and absolute percentage error rates of 1.0-12.8%. These findings underscore the importance of utilizing extensive and diverse data sets for predictive modeling to enhance accuracy and stability. Furthermore, exploring the interpretability of the model and experimentally validating newly synthesized PIs have augmented the practical utility of the model. Under the guidance of the Tg-QSPR models, it will be possible to accelerate the performance prediction and structural design of PIs in the future.

Elucidating structure-property relationships of guar gum biomolecules: insights from -polynomial and QSPR modeling.

This study investigates the quantitative structure-property relationship (QSPR) modeling of guar gum biomolecules, focusing on their structural parameters. Guar gum, a polysaccharide with diverse industrial applications, exhibits various properties such as viscosity, solubility, and emulsifying ability, which are influenced by its molecular structure. In this research, -polynomial and associated topological indices are employed as structural descriptors to represent the molecular structure of guar gum. The -polynomial and associated topological indices capture important structural features, including size, shape, branching, and connectivity. By correlating these descriptors with experimental data on guar gum properties, predictive models are developed using regression analysis techniques. The analysis revealed a strong correlation between the boiling point and molecular weight and all the considered topological descriptors. The resulting models offer insights into the relationship between guar gum structure and its properties, facilitating the optimization of guar gum production and application in various industries. This study demonstrates the utility of -polynomial and QSPR modeling in elucidating structure-property relationships of complex biomolecules like guar gum, contributing to the advancement of biomaterial science and industrial applications.

Application of machine learning models for property prediction to targeted protein degraders

Machine learning (ML) systems can model quantitative structure-property relationships (QSPR) using existing experimental data and make property predictions for new molecules. With the advent of modalities such as targeted protein degraders (TPD), the applicability of QSPR models is questioned and ML usage in TPD-centric projects remains limited. Herein, ML models are developed and evaluated for TPDs’ property predictions, including passive permeability, metabolic clearance, cytochrome P450 inhibition, plasma protein binding, and lipophilicity. Interestingly, performance on TPDs is comparable to that of other modalities. Predictions for glues and heterobifunctionals often yield lower and higher errors, respectively. For permeability, CYP3A4 inhibition, and human and rat microsomal clearance, misclassification errors into high and low risk categories are lower than 4% for glues and 15% for heterobifunctionals. For all modalities, misclassification errors range from 0.8% to 8.1%. Investigated transfer learning strategies improve predictions for heterobifunctionals. This is the first comprehensive evaluation of ML for the prediction of absorption, distribution, metabolism, and excretion (ADME) and physicochemical properties of TPD molecules, including heterobifunctional and molecular glue sub-modalities. Taken together, our investigations show that ML-based QSPR models are applicable to TPDs and support ML usage for TPDs’ design, to potentially accelerate drug discovery.

In Silico and In Vitro Studies of Terpenes from the Fabaceae Family Using the Phenotypic Screening Model against the SARS-CoV-2 Virus.

In 2019, the emergence of the seventh known coronavirus to cause severe illness in humans triggered a global effort towards the development of new drugs and vaccines for the SARS-CoV-2 virus. These efforts are still ongoing in 2024, including the present work where we conducted a ligand-based virtual screening of terpenes with potential anti-SARS-CoV-2 activity. We constructed a Quantitative Structure-Activity Relationship (QSAR) model from compounds with known activity against SARS-CoV-2 with a model accuracy of 0.71. We utilized this model to predict the activity of a series of 217 terpenes isolated from the Fabaceae family. Four compounds, predominantly triterpenoids from the lupane series, were subjected to an in vitro phenotypic screening in Vero CCL-81 cells to assess their inhibitory activity against SARS-CoV-2. The compounds which showed high rates of SARS-CoV-2 inhibition along with substantial cell viability underwent molecular docking at the SARS-CoV-2 main protease, papain-like protease, spike protein and RNA-dependent RNA polymerase. Overall, virtual screening through our QSAR model successfully identified compounds with the highest probability of activity, as validated using the in vitro study. This confirms the potential of the identified triterpenoids as promising candidates for anti-SARS-CoV-2 therapeutics.

Machine-Learning-Guided Peptide Drug Discovery: Development of GLP-1 Receptor Agonists with Improved Drug Properties.

Peptide-based drug discovery has surged with the development of peptide hormone-derived analogs for the treatment of diabetes and obesity. Machine learning (ML)-enabled quantitative structure-activity relationship (QSAR) approaches have shown great promise in small molecule drug discovery but have been less successful in peptide drug discovery due to limited data availability. We have developed a peptide drug discovery platform called streaMLine, enabling rigorous design, synthesis, screening, and ML-driven analysis of large peptide libraries. Using streaMLine, this study systematically explored secretin as a peptide backbone to generate potent, selective, and long-acting GLP-1R agonists with improved physicochemical properties. We synthesized and screened a total of 2688 peptides and applied ML-guided QSAR to identify multiple options for designing stable and potent GLP-1R agonists. One candidate, GUB021794, was profiled in vivo (S.C., 10 nmol/kg QD) and showed potent body weight loss in diet-induced obese mice and a half-life compatible with once-weekly dosing.

Prediction of Human Liver Microsome Clearance with Chirality-Focused Graph Neural Networks.

In drug candidate design, clearance is one of the most crucial pharmacokinetic parameters to consider. Recent advancements in machine learning techniques coupled with the growing accumulation of drug data have paved the way for the construction of computational models to predict drug clearance. However, concerns persist regarding the reliability of data collected from public sources, and a majority of current in silico quantitative structure-property relationship models tend to neglect the influence of molecular chirality. In this study, we meticulously examined human liver microsome (HLM) data from public databases and constructed two distinct data sets with varying HLM data quantity and quality. Two baseline models (RF and DNN) and three chirality-focused GNNs (DMPNN, TetraDMPNN, and ChIRo) were proposed, and their performance on HLM data was evaluated and compared with each other. The TetraDMPNN model, which leverages chirality from 2D structure, exhibited the best performance with a test R2 of 0.639 and a test root-mean-squared error of 0.429. The applicability domain of the model was also defined by using a molecular similarity-based method. Our research indicates that graph neural networks capable of capturing molecular chirality have significant potential for practical application and can deliver superior performance.

Photochemical transformation of liquid crystal monomers in simulated environmental media: Kinetics, mechanism, toxicity variation and QSAR modeling

Liquid crystal monomers (LCMs) are a new class of emerging pollutants with high octanol-water partition coefficients; however, their transformation behavior and associated risk to environments with high organic matter content has rarely been reported. In this study, we investigated the photodegradation kinetics, mechanism, and toxicity variation of 23 LCMs on leaf wax models (e.g., organic solvents methanol and n-hexane). The order of the photolysis rates of these LCMs were biphenylethyne LCMs > phenylbenzoate LCMs > diphenyl/terphenyl LCMs under simulated sunlight, while the phenylcyclohexane LCMs were resistant to photodegradation. The phenylbenzoate and biphenylethyne LCMs mainly undergo direct photolysis, while the diphenyl/terphenyl LCMs mainly undergo self-sensitized photolysis. The main photolysis pathways are the cleavage of ester bonds for phenylbenzoate LCMs, the addition, oxidation and cleavage of alkynyl groups for biphenylethyne LCMs, and the cleavage/oxidation of chains attached to phenyls and the benzene ring opening for diphenyl/terphenyls LCMs. Most photolysis products remained toxic to aquatic organisms to some degree. Additionally, two quantitative structure-activity relationship models for predicting kobs of LCMs in methanol and n-hexane were developed, and employed to predict kobs of 93 LCMs to fill the kobs data gap in systems mimicking leaf surfaces. These results can be helpful for evaluating the fate and risk of LCMs in environments with high content of organic phase.

Reverse Zagreb Indices and Their Application in the Evaluation of Physiochemical Properties of Anticancer/Antibacterial Drugs.

Recent advancements in chemical graph theory have facilitated a deeper understanding of the relationship between chemical structures and their underlying graphs based upon classical graph theory principles. Quantitative structure-property relationship (QSPR) analysis stands out as a powerful tool for probing chemical graphs. Topological indices, graph invariants assigning numerical values to graphs, play a pivotal role in statistically correlating physical properties, chemical reactivity, and biological activity across diverse chemical structures. The cactus graph (CG) represents a noteworthy family of connected graphs distinguished by the absence of common vertices between cycles. In this study, we focus on establishing the expressions of Zagreb and reverse Zagreb indices in terms of known parameters for cactus graphs. Subsequently, we leverage these indices to conduct QSPR analysis, employing linear and multilinear regression models to investigate the relationship between the properties of chemical structures adorned with cactus-like graphs and the corresponding Zagreb and reverse Zagreb indices.

Ecotoxicological effect factors for calculating USEtox ecotoxicity characterization factors

BackgroundThis study aims to tackle the lack of freshwater ecotoxicological effect factors (EFs) crucial for determining freshwater ecotoxicity characterization factors (CFs) using the widely accepted scientific consensus USEtox model for ecotoxicity impact characterization. The objectives are: (1) to offer a collection of experimental EFs to support USEtox ecotoxicity characterization factor computations and (2) to contrast ecotoxicity data produced by various quantitative structure–activity relationship (QSAR) models against experimental data.ResultsExperimental ecotoxicity data were gathered from the REACH database and CompTox Version 2.1.1, which includes toxicity information from ToxValDB v9.4. QSAR-driven ecotoxicity data were extracted from ECOSAR v1.11 and T.E.S.T. v5.1.2. The experimental and estimated data underwent a harmonization process to ensure consistency. Subsequently, aquatic ecotoxicological EFs were determined. The merged REACH and CompTox databases list EFs for 11,295 substances, each identified by a unique CAS number. Among these, the USEtox database already catalogs 2426 substances with freshwater ecotoxicological EFs. This study expanded on that by calculating EFs for an additional 8869 substances. Using estimated data, EFs were determined for 6029 chemicals based on ECOSAR data and 6762 chemicals using TEST data.ConclusionsThis study calculated EFs for an additional 8869 substances, thereby broadening their inclusion in LCA evaluations. When integrated with the USEtox EFs database, this research encompasses 11,368 chemicals. The high correlation observed between experimental EFs and those in the USEtox database lends significant confidence to the calculations for chemicals not listed in USEtox. Conversely, the low correlation between estimated EFs and those in USEtox suggests limited confidence in calculations based on estimated data. Furthermore, the disparity in correlations between EFs calculated using ECOSAR and TEST indicates that different QSARs can yield varied results. This discrepancy underscores the need for caution when relying on estimated data. Given that EFs are contingent on data availability, it is imperative to periodically update EFs as new data emerges.

On Certain Degree Based and Bond-additive Topological Indices of Dodeca-benzo-circumcorenene.

Chemical graph theory has been used to mathematically model the various physical and biological aspects of chemical substances. A mathematical formulation that may be applied to any graph and can characterise a molecule structure is known as a topological index or molecular descriptor. It is convenient and efficient to analyse the mathematical values and further research on various physical properties of a molecule based on these molecular descriptors. They provide useful alternatives to lengthy, expensive, and labour-intensive laboratory experiments. The topological indices can be used to predict the chemical structures, physicochemical properties, and biological activities using quantitative structure-activity relationships (QSARs) and quantitative structure-property relationships (QSPRs). In this study, the molecular descriptors of the Dodeca-benzo-circumcorenene compounds are derived based on their corresponding molecular structures. The computed indices are then compared graphically to study their relationship with the molecular structure and with each other..

Synthesis, molecular modelling studies of artemisinin-chalcone derivatives and their antimalarial activity evaluation

Twenty-two monomers and dimers of artemisinin having chalcone as a linker were synthesised, and their antimalarial activity against Plasmodium falciparum was determined, and a quantitative structure-activity relationship (QSAR) was developed. Artemisinin is a frontline antimalarial drug known worldwide but is threatened because of the rapidly emerging artemisinin-resistant strain Plasmodium falciparum. In vitro, antimalarial IC50 (half-maximal inhibitory concentration) activity of a molecule against malaria parasites provides a good first screen for identifying the antimalarial potential of a particular molecule. The most active compound was artemisinin dimer dimethoxy chalcone as a linker (22) with IC50 of 4.34 nM. The molecular mechanism was explored through in silico docking & ADMET studies for the active compounds.

Integrated pharmacophore‐based screening and docking strategy for identifying type II inhibitors of VEGFR‐2 kinase

Pharmacophore modeling and quantitative structure-activity relationship (QSAR) analysis represent indispensable methodologies in modern drug discovery, offering insights into ligand-receptor interactions and aiding in the rational design of potent therapeutic agents. This study presents a thorough investigation into the identification of novel inhibitors targeting VEGFR-2 kinase, a key regulator in angiogenesis and a promising target in oncology.Leveraging a dataset comprising 72 type-II inhibitors, a series of pharmacophore hypotheses were developed using a set of six active ligands. Subsequent evaluation led to the selection of top-performing hypotheses, ADDHRR_6 and ADDHRR_10, which demonstrated good predictive capabilities in atom-based 3D QSAR modeling.Virtual screening of the Maybridge database using these pharmacophore hypotheses resulted in the identification of potential hits, which were evaluated through molecular docking analyses. Ten compounds exhibited favorable docking scores and important interactions with key residues, such as Glu885, Glu917, Cys919, and Asp1046.Additionally, comprehensive absorption, distribution, metabolism, and excretion (ADME) analysis was performed and indicated that these compound had acceptable physicochemical properties. These findings highlight the utility of integrated pharmacophore-based screening and molecular docking approaches in the identification of novel type II VEGFR-2 kinase inhibitors.

A QSAR-based application for the prediction of lethal blood concentration of new psychoactive substances

The rapid development and introduction of new psychoactive substances (NPS) into illegal markets present an enormous challenge for forensic toxicologists, as there is limited knowledge about their toxicity in humans. To strengthen forensic interpretation of NPS intoxication cases, we have developed a predictive model for estimating human lethal blood concentrations (LBC) of various NPS. This quantitative structure-activity relationship (QSAR) model focuses on opioids, designer benzodiazepines, synthetic cathinones, synthetic cannabinoids, and phenethylamines. Utilising linear regression and multilayer perceptron algorithms, the models was trained using data from the existing literature. A toxicological significance-based approach have been applied to refine the selection of training data. The model demonstrated satisfactory performance metrics through cross-validation (R ≈ 0.8, MAE ≈ 0.6) and comparison with experimental data (R ≈ 0.9). A Python-based web application have been developed to facilite the use of the created model in predicting LBC of NPS. Despite the model's reliability, limitations due to data availability, quality and the complexities of post-mortem toxicology mean that its predictions should be interpreted with caution.

Development of an effective QSAR-based hazard threshold prediction model for the ecological risk assessment of aromatic hydrocarbon compounds.

The insufficient hazard thresholds of specific individual aromatic hydrocarbon compounds (AHCs) with diverse structures limit their ecological risk assessment. Thus, herein, quantitative structure-activity relationship (QSAR) models for estimating the hazard threshold of AHCs were developed based on the hazardous concentration for 5% of species (HC5) determined using the optimal species sensitivity distribution models and on the molecular descriptors calculated via the PADEL software and ORCA software. Results revealed that the optimal QSAR model, which involved eight descriptors, namely, Zagreb, GATS2m, VR3_Dzs, AATSC2s, GATS2c, ATSC2i, ω, and Vm, displayed excellent performance, as reflected by an optimal goodness of fit (R2adj = 0.918), robustness (Q2LOO = 0.869), and external prediction ability (Q2F1 = 0.760, Q2F2 = 0.782, and Q2F3 = 0.774). The hazard thresholds estimated using the optimal QSAR model were approximately close to the published water quality criteria developed by different countries and regions. The quantitative structure-toxicity relationship demonstrated that the molecular descriptors associated with electrophilicity and topological and electrotopological properties were important factors that affected the risks of AHCs. A new and reliable approach to estimate the hazard threshold of ecological risk assessment for various aromatic hydrocarbon pollutants was provided in this study, which can be widely popularised to similar contaminants with diverse structures.

QSAR Studies on a Series of Pyrazole Azabicyclo [3.2.1] Octane Sulfonamides N-acylethanolamine-hydrolyzing Acid Amidase Inhibitors

Background: Inflammation is a common and intractable disease for humans. Current antiinflammatory drugs have a lot of side effects, which cause irreversible damage to the body. Objective: We predict the activity of the N-acylethanolamine-hydrolyzing acid amidase (NAAA) inhibitor to find more effective compounds. Methods: We established a quantitative structure-activity relationship (QSAR) model by gene expression programming to predict the IC50 values of natural compounds. The NAAA inhibitor, as a cysteine enzyme, plays an important role in the therapy of pain, anti-inflammatory effects and application of other diseases. A total of 36 NAAA inhibitors were optimized by the heuristic method in the CODESSA program to build a linear model. The 27 compounds and 9 compounds were in train and test sets. On this basis, we selected three descriptors and used them to build nonlinear models in gene expression programming. Results: The best model in the gene expression programming method was found, the square of correlation coefficients of R2 and mean square error for the training set were 0.79 and 0.14, testing set was 0.78 and 0.20, respectively. Conclusion: From this method, the activity of molecules could be predicted, and the best method was found. Therefore, this model has a stronger predictive ability to develop NAAA inhibitors.

Exploration of carboxamide hybrid indolyl aryl sulfones as antiretroviral agents by HIV-1 reverse transcriptase inhibition and antioxidant effects: Synthesis, biochemical screening and computational analysis

Indolyl aryl sulfones (IASs) have been under investigation as HIV-1 non-nucleoside reverse transcriptase inhibitors (NNRTIs) since the discovery of Merck's compound l-737,126. This research work reports the synthesis of a novel series of indolyl aryl sulfonyl carboxamides (11 to 21 and 25 to 28) and their biological evaluation. Among the synthesized compounds, 15 (IC50 = 57.4 ± 10.1 µM) and 26 (IC50 = 34.0 ± 1.9 µM) demonstrated potent inhibitory activity in RNA-dependent DNA polymerization (RDDP) reactions. In addition, radical inhibition activity was performed by reacting the synthesized compounds against 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical and it was observed that compounds 20 (IC50 = 8.2 ± 0.2 µM) and 21 (IC50 = 9.3 ± 0.5 µM) showed promising antioxidant properties. Additionally, Two-Dimensional Quantitative Structure Activity Relationship (2D-QSAR) and molecular docking studies were employed to provide insight into the molecular mechanism of RDDP inhibition.

A Quantitative Structure–Property Relationship Model for Surface Tension Based on Artificial Neural Network

A Quantitative Structure–Property Relationship Model for Surface Tension Based on Artificial Neural Network