Quantitative Structure-activity Relationship Models Research Articles

QSAR Regression Models for Predicting HMG-CoA Reductase Inhibition

Background/Objectives: HMG-CoA reductase is an enzyme that regulates the initial stage of cholesterol synthesis, and its inhibitors are widely used in the treatment of cardiovascular diseases. Methods: We have created a set of quantitative structure-activity relationship (QSAR) models for human HMG-CoA reductase inhibitors using nested cross-validation as the primary validation method. To develop the QSAR models, we employed various machine learning regression algorithms, feature selection methods, and fingerprints or descriptor datasets. Results: We built and evaluated a total of 300 models, selecting 21 that demonstrated good performance (coefficient of determination, R2 ≥ 0.70 or concordance correlation coefficient, CCC ≥ 0.85). Six of these top-performing models met both performance criteria and were used to construct five ensemble models. We identified the descriptors most important in explaining HMG-CoA inhibition for each of the six best-performing models. We used the top models to search through over 220,000 chemical compounds from a large database (ZINC 15) for potential new inhibitors. Only a small fraction (237 out of approximately 220,000 compounds) had reliable predictions with mean pIC50 values ≥ 8 (IC50 values ≤ 10 nM). Our svm-based ensemble model predicted IC50 values < 10 nM for roughly 0.08% of the screened compounds. We have also illustrated the potential applications of these QSAR models in understanding the cholesterol-lowering activities of herbal extracts, such as those reported for an extract prepared from the Iris × germanica rhizome. Conclusions: Our QSAR models can accurately predict human HMG-CoA reductase inhibitors, having the potential to accelerate the discovery of novel cholesterol-lowering agents and may also be applied to understand the mechanisms underlying the reported cholesterol-lowering activities of herbal extracts.

Application of Machine Learning in the Development of Fourth Degree Quantitative Structure–Activity Relationship Model for Triclosan Analogs Tested against Plasmodium falciparum 3D7

Application of Machine Learning in the Development of Fourth Degree Quantitative Structure–Activity Relationship Model for Triclosan Analogs Tested against <i>Plasmodium falciparum</i> 3D7

Novel "Phenyl-Pyrazoline-Oxadiazole" Ternary Substructure Derivatives: Synthesis, Insecticidal Activities, and Structure-Activity Relationship Study.

In recent years, isoxazole insecticides or parasiticides targeting the γ-aminobutyric acid receptor, such as fluralaner or fluxametamide, featured a novel chemical structure and exhibited potent insecticidal activity with no-cross resistance. Thus, many research institutes have tried to modify the structures of these agents to find a new insecticide. Previously, the majority of researchers stuck to the "phenyl-isoxazole-phenyl" structure, making modifications only to other components. In this study, the "phenyl-isoxazole-phenyl" ternary motif was modified for the first time based on bioisosterism theory. A series of new derivatives carrying pyrazoline and 1,3,4-oxadiazole moieties were designed and synthesized to investigate their insecticidal activities against the diamondback moth (Plutella xylostella) and fall armyworm (Spodoptera frugiperda). Preliminary bioassay data showed that some of the target compounds exhibited good insecticidal activities against P. xylostella and S. frugiperda. Especially, compound A21 showed insecticidal activity against P. xylostella (LC50 = 1.2 μg/mL) better than commercial insecticide ethiprole (LC50 = 2.9 μg/mL) but worse than parasiticide fluralaner (LC50 = 0.5 μg/mL). Similarly, compound A21 exhibited insecticidal activity to S. frugiperda (LC50 = 13.2 μg/mL) better than commercial insecticide fipronil (LC50 = 78.8 μg/mL) but worse than fluralaner (LC50 = 0.7 μg/mL). Compound A21 could serve as a potential lead compound to control P. xylostella and S. frugiperda. The three-dimensional quantitative structure-activity relationship model revealed that the further introduction of an electron-donating group in the 2- or 3-site may increase the insecticidal activity of A21. Molecular dynamics simulations showed that the hydrogen bond of A21 and receptor was important for the binding receptor. This study has identified a new substructure called "phenyl-pyrroline-oxadiazole" instead of the previously known "phenyl-isoxazole-phenyl" substructure, offering a useful guide for the design of novel insecticide molecules.

Emodin derivatives as novel potent DPP-4 inhibitors: Design, synthesis, and in vitro evaluation

Dipeptidyl peptidase-4 (DPP-4) inhibitors have gained recognition as effective agents for lowering blood sugar levels, significantly improving glycemic control for individuals with type 2 diabetes mellitus (T2DM). Emodin, an anthraquinone derived from the traditional herbs rhubarb (Rheum officinale) and Polygonum cuspidatum, has been identified as an important component in the development of new treatments for diabetes. In the present work, we explored the DPP-4 inhibitory activity of emodin derivatives. This study focused on the design, synthesis, and evaluation of emodin derivatives for their in vitro DPP-4 inhibitory activity. Molecular docking studies indicated that 3-o-toluoyl emodin (OTEM) had the lowest docking score (−134.073) against the DPP-4 protein among the tested compounds. OTEM also achieved the highest drug-likeness score of 0.56 and demonstrated DPP-4 inhibitory activity, with an IC50 value of 0.77 μM. Furthermore, structure-activity relationship (SAR) analysis suggested that the addition of an ortho-toluoyl group at the C-3 position could enhance DPP-4 inhibition. Additionally, quantitative structure-activity relationship (QSAR) model assessments revealed that log P was the only descriptor significantly influencing DPP-4 inhibitory activity. Therefore, the current study indicates that OTEM could serve as a promising lead compound to address the demand for antidiabetic agents.

Computational Strategies for Assessing Adverse Outcome Pathways: Hepatic Steatosis as a Case Study.

The evolving landscape of chemical risk assessment is increasingly focused on developing tiered, mechanistically driven approaches that avoid the use of animal experiments. In this context, adverse outcome pathways have gained importance for evaluating various types of chemical-induced toxicity. Using hepatic steatosis as a case study, this review explores the use of diverse computational techniques, such as structure-activity relationship models, quantitative structure-activity relationship models, read-across methods, omics data analysis, and structure-based approaches to fill data gaps within adverse outcome pathway networks. Emphasizing the regulatory acceptance of each technique, we examine how these methodologies can be integrated to provide a comprehensive understanding of chemical toxicity. This review highlights the transformative impact of in silico techniques in toxicology, proposing guidelines for their application in evidence gathering for developing and filling data gaps in adverse outcome pathway networks. These guidelines can be applied to other cases, advancing the field of toxicological risk assessment.

Design, Synthesis, and 3D-QASR of Oxazoline Derivatives Containing an S-S Moiety as Potential Acaricide.

To mitigate the detrimental effects of agricultural pest mites on crop yield, an active substructure splicing strategy was employed to modify etoxazole by introducing an S-S moiety. The target products were obtained efficiently via a bilateral disulfurating reagent (DSMO), which was developed by our group. The leaf dip method was used to evaluate the activities of the designed target compounds against the eggs and larvae of the spider mite (Tetranychus cinnabarinus). Most of the target compounds exhibited good efficacy in controlling the larvae and eggs of T. cinnabarinus. Based on these results, a three-dimensional quantitative structure-activity relationship (3D-QSAR) model was established to guide the construction of compound 7l. Notably, compound 7l exhibited a better activity against T. cinnabarinus eggs (LC50 = 0.0035 mg/L) compared to etoxazole (LC50 = 0.2990 mg/L). Greenhouse bioassays indicated that compound 7l exhibits excellent acaricidal activity against egg of T. cinnabarinus, which is better than the etoxazole at 1.0 mg/L. Additionally, some of the compounds showed inhibitory effects against Dickeya zeae (D. zeae), Xanthomonas campestris pv campestris (Xcc), Xanthomonas oryzae pvoryza (Xoo), and Xanthomonas oryzae pvoryzicola (Xoc). Furthermore, compounds 7l not only exhibited relatively potent against Plutella xylostella activities (LC50 = 24.0 mg/L) but also had low toxicity (LC50 > 11.0 μg/bee) to Apis mellifera. In conclusion, the current experimental results suggest that oxazoline derivatives containing an S-S moiety have the potential to serve as lead compounds for the development of novel acaricide agents.

Quercetin: Potential antidiabetic effects through enzyme inhibition and starch digestibility

AbstractDiabetes mellitus involves high blood sugar levels due to insufficient insulin action. Furthermore, enzymes such as α‐amylase and α‐glucosidase break down carbohydrates into glucose, leading to postprandial hyperglycemia. Flavonoids, particularly quercetin, inhibit these enzymes, slowing carbohydrate digestion and reducing glucose absorption. Quercetin has significant hypoglycemic effects with inhibitory concentration (IC50) values comparable to acarbose, a standard inhibitor, suggesting its potential as a natural alternative for diabetes management. In silico models, including molecular docking, molecular dynamics (MD) simulations, and quantitative structure‐activity relationship (QSAR) approaches, help researchers understand the molecular interactions of therapeutic agents. These techniques identify potential inhibitors, determine enzyme‐inhibitor structures, and calculate binding energies, correlating findings with in vitro or in vivo data. Molecular docking predicts molecular orientations, MD simulations offer insights into enzyme–inhibitor dynamics, and QSAR models predict inhibitory potential based on structural properties. Studies have shown that quercetin effectively inhibits α‐glucosidase and α‐amylase by forming hydrogen bonds with specific amino acid residues. Quercetin interacts with starches and reduces their digestibility, increases the formation of resistant starch, lowers the glycemic index, and inhibits digestive enzymes. Studies show that the effects of quercetin on starch digestion vary with concentration and type of starch, and its incorporation into foods such as bakery products, pasta, etc. can significantly decrease starch hydrolysis. The incorporation of quercetin into starch matrices may aid in the development of functional foods aimed at improving glycemic control.

Cheminformatics-driven prediction of BACE-1 inhibitors: Affinity and molecular mechanism exploration

The β-secretase, sometimes referred to as BACE1 or Asp2, is the enzyme responsible for initiating the production of Aβ by breaking down the amyloid precursor protein. Hence, BACE is a pivotal target for pharmacological intervention aimed at diminishing the production of Aβ in Alzheimer's disease (AD). We did a quantitative structure-activity relationship (QSAR) study on 1235 compounds that had been experimentally reported as BACE-1 inhibitors (Ki) to find the target molecule and structural patterns linked to blocking the BACE-1 receptor. The OECD-recommended genetic algorithm-multiple linear regression (GA-MLR) QSAR model strikes a good balance between being able to make accurate predictions and understanding how things work. It achieves high values for many evaluation metrics, including R2tr = 0.8047, Q2LMO = 0.802, R2ex = 0.805, CCCex = 0.891, Q2-F1=0.801, Q2-F2=0.799, and Q2-F3=0.815. The mechanistic interpretation of QSAR has identified some nitrogen atoms that are required to block beta-secretase and make it less effective at doing its job. A positively charged aromatic carbon atom has a more significant impact on beta-secretase-1 inhibition. The docking study showed that the catalytic dye residues Asp32 and Asp228 become protonated when compound 27 is present, but they stay unprotonated when compound 27 is not present. These rings of pyrazine and pyridine interact hydrophobically with Tyr71 and Ile118 residues, confirming the pharmacophoric features seen in the QSAR data. We examined the stability of both the protein-ligand complex and the apo-protein. The apoprotein had an average gyration radius of 22.13, whereas the protein-ligand complex had an average gyration radius of 22.91. No changes were made to the results from the molecular docking, molecular dynamics (MD) simulation, MMGBSA (Molecular Mechanics Generalized Born Surface Area), or DFT analyses. Therefore, the current work could effectively contribute to the future development of BACE inhibitors in medication design.

Comparative QSAR and q-RASAR modeling for aquatic toxicity of organic chemicals to three trout species: O. Clarkii, S. Namaycush, and S. Fontinalis

Oncorhynchus clarkii, Salvelinus fontinalis, and Salvelinus namaycush are vital trout species in North America, crucial for maintaining ecological balance, economic stability, and human health. These species thrive in cold, unpolluted waters and are highly vulnerable to contaminants. Given the rapid proliferation of industrial organic chemicals, traditional in vivo toxicity testing methods are inadequate to ensure timely and comprehensive risk assessments. Therefore, we employed in silico tools, namely Quantitative Structure-Activity Relationship (QSAR) and Quantitative Read-Across Structure-Activity Relationship (q-RASAR), to efficiently predict the aquatic toxicity of chemicals. Utilizing acute median lethal concentration (LC50) data from the US EPA’s ToxValDB, we developed the first-ever species-specific QSAR and q-RASAR models. The q-RASAR models outperformed traditional QSAR models by achieving higher internal and external statistical quality for each species. Key toxicity-determining descriptors included electrotopological state indices, autocorrelation descriptors, and similarity-based RASAR descriptors. For O. clarkii, the presence of chlorine atoms and rotatable bonds significantly influenced toxicity. S. fontinalis toxicity was strongly affected by polarizability, and van der Waals volumes, while S. namaycush showed sensitivity to weak hydrogen bond acceptors and topological complexity. The models predicted the toxicity of 1172 external compounds, identifying the most and least toxic chemicals for each species. This study not only offers the first comprehensive q-RASAR models for predicting trout species-specific toxicity but also provides novel insights into species-specific toxicological modes of action. The results contribute significantly to chemical screening and prioritization in aquatic risk assessments, effectively filling critical data gaps and advancing predictive modeling techniques.

Exploring novel natural compound-based therapies for Duchenne muscular dystrophy management: insights from network pharmacology, QSAR modeling, molecular dynamics, and free energy calculations.

Muscular dystrophies encompass a heterogeneous group of rare neuromuscular diseases characterized by progressive muscle degeneration and weakness. Among these, Duchenne muscular dystrophy (DMD) stands out as one of the most severe forms. The present study employs an integrative approach combining network pharmacology, quantitative structure-activity relationship (QSAR) modeling, molecular dynamics (MD) simulations, and free energy calculations to identify potential therapeutic targets and natural compounds for DMD. Upon analyzing the GSE38417 dataset, it was found that individuals with DMD exhibited 290 upregulated differentially expressed genes (DEGs) compared to healthy controls. By utilizing gene ontology (GO) and protein-protein interaction (PPI) network analysis, this study provides insights into the functional roles of the identified DEGs, identifying ten hub genes that play a critical role in the pathology of DMD. These key genes include DMD, TTN, PLEC, DTNA, PKP2, SLC24A, FBXO32, SNTA1, SMAD3, and NOS1. Furthermore, through the use of ligand-based pharmacophore modeling and virtual screening, three natural compounds were identified as potential inhibitors. Among these, compounds 3874518 and 12314417 have demonstrated significant promise as an inhibitor of the SMAD3 protein, a crucial factor in the fibrotic and inflammatory mechanisms associated with DMD. The therapeutic potential of the compounds was further supported by molecular dynamics simulation and Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) analysis. These findings suggest that the compounds are viable candidates for experimental validation against DMD.

Identification of Novel Tyrosinase Inhibitors with Nanomolar Potency Using Virtual Screening Approaches.

Hyperpigmentation disorders are caused by excess production of the pigment melanin, catalyzed by the enzyme tyrosinase. Novel tyrosinase inhibitors are needed as therapeutic agents to treat these conditions. To discover new inhibitors, we performed a virtual screening of the ZINC20 library containing 1.4 billion compounds. An initial filter for drug-likeness, ADMET properties, and synthetic accessibility reduced the library to 10,217 hits. Quantitative structure-activity relationship (QSAR) modeling of this subset predicted nanomolar inhibitory potency for several chemical scaffolds. Comparative molecular docking studies and rigorous binding energy calculations further prioritized four cysteine-containing dipeptide compounds based on predicted strong binding affinity and mode to tyrosinase. Microsecond-long molecular dynamics simulations provided additional atomistic insights into the stability of inhibitor-enzyme binding interactions. This integrated computational workflow effectively sampled an extremely large chemical space to discover four novel tyrosinase inhibitors with half-maximal inhibitory concentration values below 10 nM. Overall, this demonstrates the power of virtual screening and multi-faceted computational techniques to accelerate the discovery of potent bioactive ligands from massive compound libraries by efficiently sampling chemical space.

Computational modeling of air pollutants for aquatic risk: Prediction of ecological toxicity and exploring structural characteristics

Assessing the aquatic toxicity originating from air pollutants is essential in sustaining water resources and maintaining the ecosystem's safety. Quantitative structure-activity relationship (QSAR) models provide a computational tool for predicting pollutant toxicity, facilitating the identification/evaluation of the contaminants and identifying responsible structural fragments. One-vs-all (OvA) QSAR is a tailored approach to address multi-class QSAR problems. The study aims to determine five distinct levels of aquatic hazard categories for airborne pollutants using OvA-QSAR modeling containing 254 air contaminants. This QSAR analysis reveals the critical descriptors of air pollutants to target for molecular modification. Various factors, including the selection of relevant mechanistic descriptors, data quality, and outliers, determine the reliability of QSAR models. By employing feature selection and outlier identification approaches, the robustness and accuracy of our QSAR models were significantly increased, leading to more reliable predictions in chemical hazard assessment. The results revealed that models using the Random Forest algorithm performed the best based on the selected descriptors, with internal and external validation accuracy ranging from 71.90% to 97.53% and 76.47%–98.03%, respectively. This study indicated that the aquatic risk of air contaminants might be attributed predominantly to their sp3/sp2 carbon ratio, hydrogen-bond acceptor capability, hydrophilicity/lipophilicity, and van der Waals volumes. These structures can be critical in developing innovative strategies to mitigate or avoid the chemicals' harmful effects. Supporting air quality improvement, this study contributes to the rapid implementation of measures to protect aquatic ecosystems affected by air pollution.

In silico models for the screening of human transthyretin disruptors

The use of New Approach Methodologies (NAMs), such as Quantitative Structure-Activity Relationship (QSAR) models, is highly recommended by international regulations to speed up hazard and risk assessment of Endocrine Disruptors, which are known to be linked to a wide spectrum of severe diseases on humans and wildlife. A very sensitive target for these chemicals is the thyroid hormone system, which plays a key role in regulating metabolic and cognitive functions. Several chemicals have been demonstrated to compete with the thyroid hormone thyroxine (T4) for binding to human thyroid hormone distributor protein transthyretin (hTTR). In this work, we generated three new datasets composed by T4-hTTR competing potencies of more than 200 heterogeneous chemicals measured by three different in vitro assays. These datasets were used for the development of new regression QSAR models. The best models were thoroughly validated by internal and external validation procedures. The mechanistic interpretation of the selected molecular descriptors provided information on structural features which are relevant to characterise hTTR binders, such as the presence of hydroxylated and halogenated aromatic rings. PCA analysis was used to rank the studied chemicals according to their increasing T4-hTTR competing potency. Hydroxylated and halogenated bicyclic aromatic compounds are ranked as the strongest hTTR binders. The new QSARs are useful to screen potential Thyroid Hormone System-Disrupting Chemicals (THSDCs), and to support the identification of sustainable alternatives to hazardous chemicals.

Application of Machine Learning Methods in Predicting Lysine-specific Histone Demethylase 1 (LSD1) Inhibitors

Background: Lysine-specific histone demethylase 1 (LSD1) is a well-known anti-cancer target for drug discovery. Novel reversible inhibitors of LSD1 are desirable to be developed. Objective: This study aimed to build reliable predictive models to evaluate the antineoplastic efficacy of compounds and reveal the structural foundation underlying the inhibitory activity of LSD1. Methods: Multiple artificial intelligence algorithms were utilized in the development of quantitative structure-activity relationship (QSAR) models. A dataset comprising 915 compounds with welldefined IC50 values against LSD1 was assembled for analysis. The structures of these compounds were described by different descriptor packages. Principal component analysis (PCA) was performed to explore the chemical space distribution of each dataset. Y-randomization test and applicability domain (AD) analysis were deployed to validate the reliability of models. Results: For regression models, a consensus model was constructed by integrating the predictions of four top-performing individual models (SVM_ECFP4, RFR_PyDescriptor, RFR_RDKIT, and TRANSNNI), resulting in enhanced predictive performance as compared to the individual models. The consensus model achieved a determination coefficient (r2, for the test set) value of 0.82. For classification models, the consensus model was derived from the amalgamation of three individual models (ASNN_RDKIT, ASNN_ECFP4, and RFR_ECFP4), and the overall prediction accuracy of the model was 0.96 for external validation. Conclusion: The reliable models could be used to identify highly active LSD1 inhibitors for the purpose of efficient virtual screening. In addition, the important molecular properties and structural fragments derived from this work can provide guidance for the structural optimization of novel LSD1 inhibitors.

Exploring Phytochemical Compounds Against Pseudomonas Aeruginosa Using QSAR, Molecular Dynamics, and Free Energy Landscape

AbstractPseudomonas aeruginosa is a versatile opportunistic bacterium that presents a considerable risk in medical environments because of its strong adaptability and resistance to multiple medications. Targeting the LasR quorum sensing system, which plays a crucial role in controlling virulence factors and biofilm formation, is a key intervention point. In this study, in silico molecular docking, machine learning‐based Quantitative Structure‐Activity Relationship (QSAR) techniques along with molecular dynamics simulation were employed to screen phytochemical compounds for their ability to inhibit the LasR QS system, a key regulator of virulence in Pseudomonas aeruginosa. This study screened 1652 phytochemicals using the ML‐based QSAR model to identify 52 phytochemicals that had better activity than the control (N‐{[3,5‐dibromo‐2‐(methoxymethoxy)phenyl]methyl}‐2‐nitrobenzamide). The in silico molecular docking approach that targeted LasR identified compounds 5281647, 57331045, and 5281672 with high binding affinity and hydrogen bonds that were comparable to the control (docking score=−10.3 kcal/mol and hydrogen bonds=4). In the 200 ns post‐molecular dynamics simulation, 5281647 exhibited a stable RMSD of 0.25 nm, which was comparable to the control. The maximum number of hydrogen bonds was exhibited by 57331045, while 5281647 and 5281672 consistently exhibited four hydrogen bonds. Overall, the Principal component analysis (PCA) and Free Energy Landscape (FEL) analyses of the complexes demonstrated that these three compounds were in stable states. In comparison to the control (ΔGTOTAL=−39.58 kcal/mol), the cumulative binding free energy (ΔGTOTAL) for 5281647 and 57331045 was −39.95 kcal/mol and −39.25 kcal/mol, respectively. This further confirms the superior binding affinity of the two compounds. Both 5281647 and 57331045 were identified as potent inhibitors of the LasR transcription factor, which is essential for the quorum sensing of Pseudomonas aeruginosa, in the present investigation. These findings underscore the importance of further exploration and optimization of phytochemicals for combating bacterial infections, offering promising avenues for future drug discovery efforts targeting this resilient pathogen.

A contemporary review of machine learning to predict adverse pregnancy outcomes from pharmaceuticals, including DDIs.

Those undergoing pregnancy are often excluded from clinical drug trials due to the risk that participation would pose. However, they often require pharmaceuticals to manage health conditions that, if gone untreated, could harm themselves or the fetus. This can mean that such individuals take one or more pharmaceuticals during pregnancy, many of which have unknown reproductive effects. Machine learning models have been used to successfully predict a number of reproductive toxicological outcomes for pharmaceuticals, including transplacental transfer, US Food and Drug Administration safety rating, and drug interactions. Models use quantitative chemical and structural features of active compounds to make predictions concerning the outcome of interest using computational algorithms. Results from these models can be a potential source of valuable information for pregnant people and their medical providers when making decisions regarding therapeutic drug use. This review summarizes current machine learning applications to make predictions about risk and toxicity of medication use during pregnancy. Our review of the recent literature revealed that machine learning quantitative structure-activity relationship models can be used successfully to predict the transplacental transfer and the US Food and Drug Administration pregnancy safety category of pharmaceuticals; such models have also been employed to predict drug interactions, though not specifically during pregnancy. This latter topic is a potential area for future research. In this review, no single algorithm or descriptor-calculation software emerged as the most widely used, and their performances depend on a variety of factors, including outcome of interest and combination of such algorithms and software.

Using the Super-learner to Predict the Chemical Acute Toxicity on Rats

With the rapid increase in the number of commercial chemicals, testing methods regarding on median lethal dose (LD50) relying animal experiments face challenges such as high costs and ethical concerns. Classical quantitative structure-activity relationship models relying on single algorithm always lack interpretability and precision, given the complexity of the mechanisms underlying acute toxicity. To address these issues, this study has developed a predictive framework using an ensemble learning model based on Super-learner. Particularly, we first obtained LD50 data for 9,843 compounds and constructed 16 meta models using 4 molecular descriptors and machine learning algorithms. The Super-learner model performed well, achieving R² values of 0.61 and 0.64 in five-fold cross-validation and test sets, respectively, with corresponding root mean square errors of 0.55 and 0.64, significantly outperforming the results of individual model. Additionally, we incorporated data filtering and applicability domain methods, which demonstrated that the Super-learner can mitigate the impact of dataset noise to some extent. The model achieved an R² of 0.76 within an applicability domain, ensuring prediction accuracy within the chemical space. Compared to previous studies, the model developed here using Super-learner generally achieved better performance across a larger applicability domain. Finally, we has launched an online tool (http://sltox.hhra.net), allowing users to quickly predict LD50 of compounds, greatly simplifying the chemical safety assessment process. This study not only provides an effective and cost-efficient method for predicting chemical toxicity but also offers technical support and data for risk assessments of chemicals.

QSAR Modeling for Predicting Beta-Secretase 1 Inhibitory Activity in Alzheimer's Disease with Support Vector Regression

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by cognitive decline, with the accumulation of β-amyloid (Aβ) plaques playing a key role in its progression. Beta-Secretase 1 (BACE1) is a crucial enzyme in Aβ production, making it a prime therapeutic target for AD treatment. However, designing effective BACE1 inhibitors has been challenging due to poor selectivity and limited blood-brain barrier permeability. To address these challenges, we employed a machine learning approach using Support Vector Regression (SVR) in a Quantitative Structure-Activity Relationship (QSAR) model to predict the inhibitory activity of potential BACE1 inhibitors. Our model, trained on a dataset of 7,298 compounds from the ChEMBL database, accurately predicted pIC50 values using molecular descriptors, achieving an R² of 0.690 on the testing set. The model's performance demonstrates its utility in prioritizing drug candidates, potentially accelerating drug discovery. This study highlights the effectiveness of computational approaches in optimizing drug discovery and suggests that further refinement could enhance the model’s predictive power for AD therapeutics.

Photodegradation of polychlorinated biphenyls (PCBs) on suspended particles from the Yellow River under sunlight irradiation: QSAR model and mechanism analysis

Polychlorinated biphenyls (PCBs), as a class of hydrophobic organic pollutants, are widely found in river sediments and suspended particles, and the environmental fate of different PCBs can be better understood by investigating their photochemical transformation process. In this study, the quantitative structure-activity relationship (QSAR) model between the photodegradation rate constants of 17 PCBs adsorbed on Yellow River suspended particles in water (as a typical heterogeneous photodegradation system) and the physicochemical parameters of PCBs was constructed by SPSS and machine learning. The model showed that the more hydrophobicity of the molecule, the more positive charge carried by the aromatic C atoms, and the presence of chlorine atoms adjacent to the carbon bridge could all enhance the photochemical activity of PCBs. From the combined analysis of rate constants, quenching experiments and theoretical calculations, it was revealed for the first time that in natural suspended particle containing organic matter, the higher concentration of •O2- and 1O2 in the hydrophobic zone contributed more to the more hydrophobic PCBs, while •OH in the hydrophilic zone played a major role in the degradation of the less hydrophobic PCBs. Findings of this study would deepen the understanding of the degradation mechanism of hydrophobic pollutants by active species in complex environments.

QSAR: Using the Past to Study the Present.

Quantitative structure-activity relationships (QSAR) is a method for predicting the physical and biological properties of small molecules; it is in use in industry and public services. However, as any scientific method, it is challenged by more and more requests, especially considering its possible role in assessing the safety of new chemicals. To answer the question whether QSAR, by exploiting available knowledge, can build new knowledge, the chapter reviews QSAR methods in search of a QSAR epistemology. QSAR stands on tree pillars, i.e., biological data, chemical knowledge, and modeling algorithms. Usually the biological data, resulting from good experimental practice, are taken as a true picture of the world; chemical knowledge has scientific bases; so if a QSAR model is not working, blame modeling. The role of modeling in developing scientific theories, and in producing knowledge, is so analyzed. QSAR is a mature technology and is part of a large body of in silico methods and other computational methods. The active debate about the acceptability of the QSAR models, about the way to communicate them, and the explanation to provide accompanies the development of today QSAR models. An example about predicting possible endocrine-disrupting chemicals (EDC) shows the many faces of modern QSAR methods.