Relevance. Modern artificial neural network models require significant energy and other resources for training and operation. Training generative models involves vast amounts of data. At the same time, these models face challenges related to the trustworthiness of the information they generate. An alternative to current paradigms of building and training neural networks is the development of energy-based models, which could potentially overcome these shortcomings and bring information processing closer to biologically and physically grounded processes. However, existing energy-based models differ little from classical models in terms of their limitations and drawbacks. Therefore, developing new approaches to modeling energy-based information processing in neural networks is highly relevant. The object of research is the process of information processing in artificial neural networks. The subject of the research is the mathematical models for the construction and training of artificial neural networks. The purpose of this paper is to develop and experimentally validate a theoretical framework that postulates the energetic nature of information and its role in the self-organization and evolution of complex information systems. Research Results. A fundamental theory is proposed, describing information as a structure of perceived external energy parameters that govern the processes of forming the internal energetic structure of a system—its model of the external world. This theory encompasses concepts of energy landscapes, principles of energy-based structural and parametric reduction, and a critical analysis of existing computational paradigms. Experimental studies on the construction and training of the developed energy-based model confirm its high generalization ability in one-pass training without using the backpropagation algorithm on ultra-small training datasets.

Reinforced concrete is a widely used construction material in various structures, predicting shear capacity in reinforced concrete beams a critical aspect of structural design. Conventional methods, such as ACI 318, often have limitations in capturing complex relationships among design variables, particularly in non-standard conditions. This study aims to develop a predictive model for the shear capacity of reinforced concrete beams using Artificial Neural Networks (ANN). Experimental data encompassing geometric, material, and load parameters were collected from the literature to construct the dataset. The development process involved designing the ANN architecture, training the model, and validating it using performance metrics such as Mean Squared Error (MSE), Coefficient of Determination (R²), and Mean Absolute Percentage Error (MAPE). The results indicate that the ANN model provides higher predictive accuracy compared to conventional methods, with a superior ability to capture complex variable relationships. Furthermore, this study offers practical implementation guidelines for utilizing the ANN model in predicting the shear capacity of reinforced concrete beams. The contributions of this research are expected to support the development of AI-based predictive methods in civil engineering, enhance design accuracy, and promote the adoption of modern technologies in civil engineering projects.

Abstract Marine ecosystems, particularly in the Gulf of Aqaba, face growing threats from anthropogenic activities and industrial pollution, necessitating advanced monitoring for sustainable management. This study presents a real-time remote monitoring (RTRM) system that integrates wireless sensor networks and machine learning (ML) to enhance water quality assessment. By continuously capturing key parameters, dissolved oxygen (DO), pH, conductivity, turbidity, and sediment concentration, the system enables dynamic tracking of ecological health. Biosensors and physicochemical sensors were combined with geographic information system-based spatial analysis. At the same time, random forest and artificial neural network models were trained on 6 months of data and validated through 10-fold cross-validation, DO: RMSE = 0.45 mg/L, R 2 = 0.92. The RTRM system provided automated analytics and early warnings for environmental risks, including coral bleaching and pollutant spills. Results showed that DO levels generally supported aquatic life, though northern coastal areas were more vulnerable due to localized pollution. Turbidity and sediment patterns highlighted recreational disturbances, particularly from boating. Compared to traditional methods, the RTRM system improved predictive accuracy by 20% and reduced monitoring costs by 30%. By unifying in situ sensing, remote sensing, and ML-based forecasting, this framework offers a scalable, cost-effective tool for real-time marine ecosystem management in the Red Sea and comparable regions.

PurposeThe integration of artificial intelligence (AI) with hyperspectral imaging (HSI) offers a promising avenue for improving pre-therapeutic prognosis, a key factor in optimizing cancer treatment strategies. This study explores the potential of artificial neural networks (ANNs) to predict the effectiveness of preoperative chemo- or radiochemotherapy in esophageal adenocarcinoma (EAC), using HSI data derived from histopathological tissue samples.MethodsHSI data were obtained from pre-therapeutic histopathological samples of 21 patients with EAC. Following annotation and spectral extraction, the data underwent pre-processing steps including normalization, shuffling, and batch organization. Three artificial neural network (ANN) models—2D convolutional neural networks (2D-CNNs), 3D convolutional neural networks (3D-CNNs), and Hybrid-Spectral Networks (Hybrid-SN)—were trained to predict treatment response. Model performance was assessed using sensitivity, specificity, accuracy, and F1-score, offering insights into their clinical utilityResultsThe 3D-CNN model achieved the highest accuracy (0.68 ± 0.09) and F1-score (0.66 ± 0.08), highlighting its strength in capturing both spatial and spectral information. The Hybrid-SN model demonstrated the highest sensitivity (0.79 ± 0.19), indicating strong performance in identifying responders to neoadjuvant therapy. In contrast, the 2D-CNN model achieved the highest specificity (0.73 ± 0.15), reflecting its effectiveness in correctly identifying non-responders.ConclusionThis study demonstrates the potential of combining HSI with ANNs to predict treatment response in EAC. Among the models evaluated, the 3D-CNN showed the most balanced performance, effectively leveraging spatial and spectral features, while the Hybrid-SN and 2D-CNN models excelled in sensitivity and specificity, respectively. These findings underline the feasibility of using AI-driven analysis of histopathological HSI data to support personalized treatment planning in EAC, paving the way for more accurate and tailored therapeutic strategies.

This study develops an optimized Artificial Neural Network (ANN) model with 7 hidden layers, TanH activation function, and 30 iterations to enhance predictive modeling for nonlinear data analysis in propulsion-related computational applications. Utilizing a dataset of 2,457 records, normalized via the Min-Max method, the model predicts key performance parameters with high accuracy, achieving R²=0.9864, RMSE=0.0110, and MAD=0.004849922 on the validation set. Compared to benchmark methods, the ANN outperforms Bootstrap Forest (R²=0.9793, RMSE=0.0136, MAD=0.005890918) and Linear Regression (R²=0.6334, RMSE=0.0574, MAD=0.031582035). Significant input variables, such as normalized operational conditions (p<0.0001) and system configuration (p<0.0001), drive the model’s performance, supporting efficient computational analysis. These findings provide a robust tool for optimizing propulsion system design, contributing to advancements in computational methods for aerospace applications.

Objective: To develop a method for predicting the risk of primary and recurrent anterior abdominal wall hernias based on the analysis of currently available diagnostic tools for assessing connective tissue status in patients. Material and methods. A comprehensive study of genome-wide association study (GWAS)-significant genetic markers, skin autofluorescence indices, and morphological changes in the skin was conducted in a cohort of 577 patients, including 299 individuals with anterior abdominal wall hernias and 278 without hernias or clinical signs of connective tissue dysplasia. An artificial neural network model was developed based on the collected data to determine the relative significance of the investigated etiopathogenetic factors. Additionally, a correlation analysis of independent variables was performed to identify associations with the diagnosis of anterior abdominal wall hernia. Binary logistic regression was used to calculate the probability of hernia occurrence or recurrence based on the identified variables. Results. Patients with anterior abdominal wall hernias demonstrated significantly lower collagen type I to type III ratios and decreased skin autofluorescence indices. Genotypic association analysis revealed a statistically significant link between the rs2009262 polymorphism of the EFEMP1 gene and increased hernia risk (p=0.033). Further evaluation using neural network analysis and correlation studies identified three independent and statistically significant predictors for hernia development: the rs2009262 EFEMP1 polymorphism, skin autofluorescence index, and collagen type I/III ratio. Conclusion: The developed predictive approach allows for individualized assessment of anterior abdominal wall hernia risk and the likelihood of recurrence in each patient.

Accurate short-term forecasting of urban water demand is a persistent challenge for utilities seeking to optimize operations, reduce energy costs, and enhance resilience in smart distribution systems. This study presents a multi-scale Artificial Neural Network (ANN) modeling approach that integrates temporal lag optimization to predict daily and hourly water consumption across heterogeneous user profiles. Using high-resolution smart metering data from the SunRise Smart City Project in Lille, France, four demand nodes were analyzed: a District Metered Area (DMA), a student residence, a university restaurant, and an engineering school. Results demonstrate that incorporating lagged consumption variables substantially improves prediction accuracy, with daily R2 values increasing from 0.490 to 0.827 at the DMA and from 0.420 to 0.806 at the student residence. At the hourly scale, the 1-h lag model consistently outperformed other configurations, achieving R2 up to 0.944 at the DMA, thus capturing both peak and off-peak consumption dynamics. The findings confirm that short-term autocorrelation is a dominant driver of demand variability, and that ANN-based forecasting enhanced by temporal lag features provides a robust, computationally efficient tool for real-time water network management. Beyond improving forecasting performance, the proposed methodology supports operational applications such as leakage detection, anomaly identification, and demand-responsive planning, contributing to more sustainable and resilient urban water systems.

Machine learning (ML) algorithms have promoted the development of predictive modeling of mineral prospectivity, enabling data-driven decision-making processes by integrating multi-source geological information, leading to efficient and accurate prediction of mineral exploration targets. However, it is challenging to conduct ML-based mineral prospectivity mapping (MPM) in under-explored areas where scarce data are available. In this study, the Narigongma district of the Qiangtang block in the Himalayan–Tibetan orogen was chosen as a case study. Five typical alterations related to porphyry mineralization in the study area, namely pyritization, sericitization, silicification, chloritization and propylitization, were extracted by remote sensing interpretation to enrich the data source for MPM. The extracted alteration evidences, combined with geological, geophysical and geochemical multi-source information, were employed to train the ML models. Four machine learning models, including artificial neural network (ANN), random forest (RF), support vector machine and logistic regression, were employed to map the Cu-polymetallic prospectivity in the study area. The predictive performances of the models were evaluated through confusion matrix-based indices and success-rate curves. The results show that the classification accuracy of the four models all exceed 85%, among which the ANN model achieves the highest accuracy of 96.43% and a leading Kappa value of 92.86%. In terms of predictive efficiency, the RF model outperforms the other models, which captures 75% of the mineralization sites within only 3.5% of the predicted area. A total of eight exploration targets were delineated upon a comprehensive assessment of all ML models, and these targets were further ranked based on the verification of high-resolution geochemical anomalies and evaluation of the transportation condition. The interpretability analyses emphasize the key roles of spatial proxies of porphyry intrusions and geochemical exploration in model prediction as well as significant influences everted by pyritization and chloritization, which accords well with the established knowledge about porphyry mineral systems in the study area. The findings of this study provide a robust ML-based framework for the exploration targeting in greenfield areas with good outcrops but low exploration extent, where fusion of a remote sensing technique and multi-source geo-information serve as an effective exploration strategy.

Fused filament fabrication (FFF) is a widely adopted 3D printing technique for the rapid, cost-effective, and customized fabrication of complex microfluidic channels using polylactic acid (PLA), particularly for drug delivery and biomedical applications. However, achieving optimal surface finish and minimizing volatile organic compound (VOC) emissions during printing remains a substantial challenge. The current study proposes a new approach by integrating a hybrid framework of artificial neural networks (ANN) with the non-dominated sorting genetic algorithm II (NSGA-II) to enhance both the quality and environmental safety of the FFF 3D printing process for microfluidic channel applications. The proposed approach optimizes four key process parameters such as layer thickness (LT), print speed (PS), material flow rate (MFR), and raster angle (RA) with a prime objective of obtaining an excellent surface finish with minimum VOC emissions. A customized FFF 3D printer embedded with cost-effective, high-sensitivity emission sensors was designed and developed to facilitate real-time monitoring, thereby offering a unique and economical capability not available in conventional commercial printers. Experimental data generated using central composite design (CCD) were employed to train a high-fidelity ANN model (R² > 95%), which functions as a surrogate model for NSGA-II-based multi-objective optimization. The ANN model demonstrated strong predictive accuracy with low mean squared error (MSE) and high correlation coefficients (R² = 0.9967 for training, 0.956 for validation, and 0.9261 for testing). The proposed ANN-NSGA-II framework identified Pareto-optimal solutions at LT (0.15 mm), PS (40 mm/s), MFR (100%), and RA (30º). The optimization results demonstrate effective control of process parameters, yielding the dual benefit of enhanced surface finish and reduced VOC emissions. The novelty of this work lies in integration of real-time emission sensing with predictive intelligence and evolutionary multi-objective optimization, ensuring high print quality while simultaneously advancing environmentally responsible additive manufacturing (AM) practices.

Structural design often neglects the dynamic effects induced by human activities. Excessive vibrations in structures such as pedestrian bridges, grandstands, slabs, and stairways have highlighted the analysis as dynamic systems of humans interacting with structures. This phenomenon, commonly referred to as “human–structure interaction” (HSI), is investigated in this study using experimental records obtained from a cantilever steel frame specially constructed to represent a variety of structures susceptible to the HSI phenomenon. This study aims to develop and evaluate artificial neural network (ANN) models capable of representing subjects in the passive condition of HSI using only simple anthropometric parameters. Two models—Nonlinear Auto-Regressive with eXogenous input (NARX) and MultiLayer Perceptron (MLP) —are implemented and compared with a conventional Mass-Spring-Damper (MSD) model. The results show that the ANN models significantly outperform the MSD model, achieving lower Normalized Mean Square Error (NMSE) values both in time-response prediction (20.23% for NARX and 25.07% for MLP vs. 30.19% for MSD) and frequency-response prediction (16.00% for NARX and 17.05% for MLP vs. 26.01% for MSD). These findings demonstrate that the proposed ANN-based models can predict the dynamic response of individual subjects using only simple anthropometric parameters such as mass and height. This approach provides a practical and efficient tool for modeling HSI in civil engineering applications.

ABSTRACT The large-scale use of concrete requires reliable quality assessment to ensure workability, mechanical properties, and durability. Conventional testing methods are often costly and time-consuming. This study explores predictive modeling as an efficient alternative, focusing on self-compacting concrete (SCC) produced and cured with seawater, silica fume, and fly ash. Workability indicators, including slump flow, J-ring, visual stability index (VSI), and air content, were used to predict compressive strength and chloride concentration. Artificial neural networks (ANNs) and Classification and Regression Trees (CART) were applied. The ANN models achieved high accuracy, with compressive strength predicted at a minimum mean squared error (MSE) of 0.085638. The chloride content prediction achieved an R² of 0.9429. CART analysis revealed that air content was the most significant factor influencing compressive strength, while the J-ring had the strongest impact on chloride content. A comparative study demonstrated that ANNs outperformed random forest regression in predictive capability. These results highlight the value of machine learning in concrete research, offering a cost-effective and time-saving method for property evaluation. The findings also support the sustainable use of seawater and supplementary cementitious materials in the production of concrete. The novelty of this study lies in predicting the compressive strength and chloride ion concentration of self-compacting concrete produced and cured with seawater and pozzolans. Neural networks and machine learning were applied for this prediction, an approach not previously explored.

ABSTRACTPrecise estimation of rice nitrogen (N) content is essential for optimizing fertilizer use. Traditional methods for estimating N content are time‐consuming, laborious, and costly. Unmanned aerial vehicles (UAVs) are time and money efficient substitutes allowing more accurate and flexible monitoring for larger rice areas. The objectives of this study were to: (i) develop random forest (RF) and artificial neural network (ANN) models for predicting and mapping the nitrogen content (%) in rice using seven vegetation indices derived from UAV multispectral sensors and; (ii) assess the key vegetation indices (VI) and their interrelationships with the predicted nitrogen content. Experiments were conducted at two locations in Cuttack district of Odisha, India, with different nitrogen levels. The UAV images were collected synchronizing with the maximum tillering stage of rice and seven indices were generated. The rice sampling was done on the date of flying UAV images and nitrogen content was estimated in the laboratory. RF and ANN models were developed using the N content as dependent and the VIs as independent variables. Both the models exhibited robust predictive capabilities, however, the RF model exhibited better performance, compared to the ANN model. Nitrogen content prediction using the developed RF and ANN models in testing site at farmer's field ranged from 0.78% to 1.95% (R2 of 0.67%) and from 0.5% to 1.78% (R2 of 0.55%), respectively. Normalized difference red edge (NDRE) and normalized difference vegetation index (NDVI) turned out as significant contributors in the development of both the models.

Deep learning-enhanced multi-modal modeling for electrosorption performance prediction via Nyquist plots.

Financial time series forecast is challenging due to nonlinear mean dynamics, volatility clustering, and long-memory effects. Traditional hybrid models such as Autoregressive Integrated Moving Average – Generalised Autoregressive Conditional Heteroscedasticity (ARIMA–GARCH) and Fractional Generalised Integrated Autoregressive Conditional Heteroscedasticity – Artificial Neural Network (FIGARCH–ANN), improve forecasting performance but remain limited by linear mean assumptions, short-memory volatility, or incomplete treatment of nonlinearities. These constraints are particularly evident in emerging markets like Nigeria, where financial returns display pronounced nonlinear and persistent volatility patterns. Thus, this study developed a hybrid model to address volatility, nonlinearity, and long memory in residuals. Daily Nigeria All Share Stock Index Data (2001-2019), exhibiting these characteristics was used to assess the forecast performance of the new hybrid Exponential Autoregressive – Fractional Integrated Generalized Autoregressive Conditional Heteroscedasticity – Artificial Neural Network (ExpAR-FIGARCH-ANN) model in comparison to the existing Exponential Autoregressive – Fractional Integrated Generalized Autoregressive Conditional Heteroscedasticity (ExpAR-FIGARCH) and Artificial Neural Network (ANN) models using error-based metrics, viz Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE) and Mean Squared Error (MSE). The empirical findings show that the hybrid ExpAR-FIGARCH-ANN model outperformed the standalone ExpAR-FIGARCH and ANN model. It achieved the lowest error metrics (MSE = 0.0029, MAE = 0.0352, MAPE = 1.68%), confirming superior predictive performance. This enhanced performance is ascribed to the novel capability of the model to concurrently address nonlinear mean dynamics, long-memory volatility, and residual nonlinearities. It provides a more accurate forecast than existing hybrid models, thus, has potential applications beyond stock indices.

ABSTRACTHydropower development significantly impacts the fragile fish habitats in river reaches of the Tibetan Plateau. To support the conservation of fish resources in these reaches, this study developed a physical habitat evaluation model for spawning grounds based on the ecological requirements of key fish species. An artificial neural network (ANN) model was employed to fit the response relationships between spawning ground indicators and environmental factors. Results indicate that water temperature is a critical factor influencing spawning grounds. In natural river reaches, suitable spawning periods occur primarily in the afternoon. In contrast, water temperature in dam‐downstream reaches is significantly affected by hydropower operations, leading to distinct differences in spawning rhythms compared to natural reaches. The Weighted Usable Area (WUA) and Patch Number (PN) of spawning grounds initially increase and then decrease with rising flow. The ANN model effectively fits the response relationships between environmental factors and WUA and PN (R2 > 0.87). Water temperature exhibits a stronger influence, while flow primarily affects WUA and PN by altering suitable substrate area. This study presents the development and application of physical and ANN models for fish spawning grounds in hydropower‐affected river reaches of the Tibetan Plateau. The findings reveal the distribution patterns of spawning grounds and identify key environmental factors. These results provide methodological references and scientific evidence for the evaluation and conservation of fish resources, supporting the sustainable management of native fish populations in plateau rivers.

Ionizing radiation such as X-rays generates reactive oxygen species (ROS), leading to oxidative stress and damage to biomolecules including hemoglobin and plasma proteins. This study aimed to evaluate the protective effects of the aqueous extract of Myrtus communis L. leaves against oxidative alterations caused by X-ray exposure. Twenty-four adult male Wistar rats were randomly assigned to control, X-ray exposed, and extract-treated plus X-ray exposed groups. The Myrtus communis extract (0.5 mg/kg) was administered intraperitoneally for 7 consecutive days. Rats in the experimental group were exposed to 6 MV X-ray radiation, and blood samples were collected one hour post-exposure. Oxidative modifications of hemoglobin (Hb) were analyzed, and plasma oxidative stress markers including protein carbonyl (PCO), malondialdehyde (MDA), and ferric reducing antioxidant power (FRAP) were measured. Artificial neural network (ANN) models were developed to identify key predictors of oxyhemoglobin (OxyHb) concentration. X-ray exposure significantly increased levels of methemoglobin (metHb) and hemichrome (HMC), while reducing absorbance at 340 nm (globin-heme interaction), 420 nm (Soret band), 542 nm (OxyHb), and 577 nm (heme-heme interaction). Plasma MDA and PCO levels were also significantly elevated. MDA showed a negative correlation with OxyHb and a positive correlation with both metHb and HMC concentrations. Administration of Myrtus communis extract effectively mitigated these oxidative changes. ANN analysis revealed that absorbance at 577 and 560 nm, metHb levels, the A577/A542 ratio, and HMC were the strongest predictors of OxyHb concentration. The aqueous extract of Myrtus communis L. leaves offers significant protection against X-ray-induced oxidative damage to hemoglobin and plasma. Moreover, ANN models can effectively identify biomarkers associated with oxidative stress in irradiated blood, with potential clinical implications for radiotherapy patients.

Nanofluids have recently emerged as a significant working fluid in thermal management and fluid dynamics. Precise assessment of their thermophysical properties, particularly thermal conductivity, is essential for evaluating heat transfer efficiency. This paper introduces an innovative statistical approach for optimizing and accurately forecasting the thermal conductivity of Cu–H 2 O nanofluid via response surface methodology (RSM). An artificial neural network (ANN) was utilized to estimate the thermal conductivity of Cu–H 2 O nanofluid. To develop the proposed thermal conductivity model, 20 experimental data points were employed, using the volume % of the nanoparticle, temperature, and diameter of the copper nanoparticle as input variables in the models. The anticipated effectiveness of the established RSM and ANN techniques is validated by experimental data and existing empirical correlations. Various statistical metrics, including margin of deviation (MOD), mean squared error (MSE), root mean squared error (RMSE), and coefficient of determination ( R 2 ), have been utilized for the quantitative evaluation of the accuracy of the developed RSM and ANN models. The consequences affirm that the created ANN models serve as a viable instrument for analyzing the behavior of industrial applications utilizing nanofluids as the operating fluid. The consequences indicated that the developed model effectively predicts the thermal conductivity of Cu–H 2 O nanofluid throughout a broad spectrum of temperature, volume percentage, and nanoparticle diameter. The constructed ANN model demonstrated optimal fit and acceptable concordance with the experimental data. The suggested correlation demonstrates superior accuracy relative to all prior correlations for predicting thermal conductivity under varying conditions.

Machine learning framework to predict product distribution of lignocellulosic biomass pyrolysis.

This work addresses the three-dimensional flow and heat transfer characteristics of nanolayered water-based nanofluid across a stretching sheet. The Boger fluid model is integrated with the Powell–Eyring fluid model to reduce the impacts of inertial forces. A high-viscosity fluid is mixed with water (base fluid) during the mixture preparation. The Koo–Kleinstreuer–Li correlation is deployed, which accounts for effective viscosity and thermal conductivity. The non-Fourier thermal relaxation effects are captured via the Cattaneo–Christov heat flux model. The governing equations are derived under the Oberbeck–Boussinesq approximation. The obtained equations are converted into dimensionless form and solved numerically using the three-stage Lobatto method. A robust Levenberg–Marquardt (LM) backpropagated artificial neural network (ANN) is trained to forecast flow dynamics. The numerical dataset is split in such a way that 15% is used for testing, 15% for validation, and 70% for training. Regression analysis, surface stresses, error histogram, correlation index, heat transfer, and MSE-based fitness curves, which range from 10−10 to 10−8 are used to validate the consistency and efficacy of LM-ANN. The findings suggest that Marangoni convection improves the axial and transverse velocities. The temperature and transverse velocity are increasing functions of the Powell–Eyring viscosity parameter, while a decreasing trend was seen for axial velocity. The temperature profile is suppressed by the thermal relaxation parameter. The ANN predicted R2 values for skin friction and Nusselt number are above 99%, which is in good agreement with the numerical results. The ANN model provides enhanced predictive ability with less processing load than traditional methods for modeling fluid dynamics.

In the search for new and effective anticancer agents, we performed a QSAR study on a series of sixteen bastadins to evaluate their potential as ACAT1 inhibitors and predict their antiangiogenic activity. Our goal was to establish a clear correlation between their biological responses and a set of molecular descriptors, applying principal component analysis (PCA), multiple linear regression (MLR), multiple nonlinear regression (MNLR), and an artificial neural network (ANN). Among the models generated, the best MLR and MNLR approaches achieved determination coefficients (R²) of 0.71 and 0.91. To further assess their reliability, we performed an external validation on a test set of three compounds, confirmed their predictive accuracy, and yielded R² test values of 0.70 and 0.83, respectively. Furthermore, the ANN model, built with a 4-4-1 architecture, showed excellent performance, achieving a correlation coefficient of 0.96 with leave-one-out cross-validation coefficients (Q²) of 0.79. These results indicate that the selected descriptors and calculated parameters are sufficient to reliably predict the biological activity of bastadins as ACAT1 inhibitors, providing a solid basis for the computer-aided design of novel anticancer agents.