Computational Model Research Articles

Cell Cycle Complexity: Exploring the Structure of Persistent Subsystems in 414 Models.

Background: The regulation of cellular proliferation and genomic integrity is controlled by complex surveillance mechanisms known as cell cycle checkpoints. Disruptions in these checkpoints can lead to developmental defects and tumorigenesis. Methods: To better understand these mechanisms, computational modeling has been employed, resulting in a dataset of 414 mathematical models in the BioModels database. These models vary significantly in detail and simulated processes, necessitating a robust analytical approach. Results: In this study, we apply the chemical organization theory (COT) to these models to gain insights into their dynamic behaviors. COT, which handles both ordinary and partial differential equations (ODEs and PDEs), is utilized to analyze the compartmentalized structures of these models. COT's framework allows for the examination of persistent subsystems within these models, even when detailed kinetic parameters are unavailable. By computing and analyzing the lattice of organizations, we can compare and rank models based on their structural features and dynamic behavior. Conclusions: Our application of the COT reveals that models with compartmentalized organizations exhibit distinctive structural features that facilitate the understanding of phenomena such as periodicity in the cell cycle. This approach provides valuable insights into the dynamics of cell cycle control mechanisms, refining existing models and potentially guiding future research in this area.

Human motor learning dynamics in high-dimensional tasks.

Conventional approaches to enhance movement coordination, such as providing instructions and visual feedback, are often inadequate in complex motor tasks with multiple degrees of freedom (DoFs). To effectively address coordination deficits in such complex motor systems, it becomes imperative to develop interventions grounded in a model of human motor learning; however, modeling such learning processes is challenging due to the large DoFs. In this paper, we present a computational motor learning model that leverages the concept of motor synergies to extract low-dimensional learning representations in the high-dimensional motor space and the internal model theory of motor control to capture both fast and slow motor learning processes. We establish the model's convergence properties and validate it using data from a target capture game played by human participants. We study the influence of model parameters on several motor learning trade-offs such as speed-accuracy, exploration-exploitation, satisficing, and flexibility-performance, and show that the human motor learning system tunes these parameters to optimize learning and various output performance metrics.

Research on the Impact of Geographical Environment on English Learning Based on Machine Learning

College English course is an important basic course in colleges and universities. The purpose of this study is to better understand the influence mechanism of geographical environment on English learning and provide scientific basis for formulating targeted English education policies. This study uses the method of machine learning to analyze the collected data and explore the influence of geographical environment on college students' English learning. This study constructs a comprehensive data set including students' English learning achievements, academic background and geographical location, and obtains the results through computer model analysis. The results show that geographical environment has an influence on college English learning, which will lead to differences in students' English learning achievements. This study suggests that geographical factors should be considered when formulating English education policies. This study is of great significance to promoting the development of English education and improving students' learning effect.

Computational study of micromagnetorotation on the micropolar flow characteristics in a rectangular channel with continuous applied shear

The aim of this investigation is to computationally analyse the effect of micromagnetorotation on flow characteristics within magneto-micropolar flows. To the authors' knowledge such consideration has not been discussed in the literature in present context. The assumption in literature that the magnetization is parallel to the applied magnetic field is incorrect due to the anisotropic nature of micropolar flows. This allows to re-investigate the micropolar flows in the present of micromagnetorotation. To this end, the theory of magneto-micropolar flows is briefly described and a computational model governing the system dynamics is presented. Finite element code in FreeFem++ is developed and validated through an exact solution. It turns out that the numerical findings obtained are in strong agreement with the analytical results. The results obtained are significant because they can assist researchers in allied fields of electronics, materials science and biomedical engineering in finding solutions to a variety of real-world issues.

Explainable artificial intelligence (XAI) to find optimal in-silico biomarkers for cardiac drug toxicity evaluation

The Comprehensive In-vitro Proarrhythmia Assay (CiPA) initiative aims to refine the assessment of drug-induced torsades de pointes (TdP) risk, utilizing computational models to predict cardiac drug toxicity. Despite advancements in machine learning applications for this purpose, the specific contribution of in-silico biomarkers to toxicity risk levels has yet to be thoroughly elucidated. This study addresses this gap by implementing explainable artificial intelligence (XAI) to illuminate the impact of individual biomarkers in drug toxicity prediction. We employed the Markov chain Monte Carlo method to generate a detailed dataset for 28 drugs, from which twelve in-silico biomarkers of 12 drugs were computed to train various machine learning models, including Artificial Neural Networks (ANN), Support Vector Machines (SVM), Random Forests (RF), XGBoost, K-Nearest Neighbors (KNN), and Radial Basis Function (RBF) networks. Our study’s innovation is leveraging XAI, mainly through the SHAP (SHapley Additive exPlanations) method, to dissect and quantify the contributions of biomarkers across these models. Furthermore, the model performance was evaluated using the test set from 16 drugs. We found that the ANN model coupled with the eleven most influential in-silico biomarkers namely dVmdtrepol,dVmdtmax,APD90,APD50,APDtri,CaD90,CaD50,Catri,CaDiastole,qInward,andqNet\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\frac{dVm}{dt}_{repol}, \\frac{dVm}{dt}_{max}, {APD}_{90}, {APD}_{50}, {APD}_{tri}, {CaD}_{90}, {CaD}_{50}, {Ca}_{tri}, {Ca}_{Diastole}, qInward, and qNet$$\\end{document} showed the highest classification performance among all classifiers with Area Under the Curve (AUC) scores of 0.92 for predicting high-risk, 0.83 for intermediate-risk, and 0.98 for low-risk drugs. We also found that the optimal in silico biomarkers selected based on SHAP analysis may be different for various classification models. However, we also found that the biomarker selection only sometimes improved the performance; therefore, evaluating various classifiers is still essential to obtain the desired classification performance. Our proposed method could provide a systematic way to assess the best classifier with the optimal in-silico biomarkers for predicting the TdP risk of drugs, thereby advancing the field of cardiac safety evaluations.

Efficient Electrocatalytic Hydrodechlorination and Detoxification of Chlorophenols by Palladium-Palladium Oxide Heterostructure.

Electrocatalytic hydrodechlorination is a promising approach for simultaneous pollutant purification and valorization. However, the lack of electrocatalysts with high catalytic activity and selectivity limits its application. Here, we propose a palladium-palladium oxide (Pd-PdO) heterostructure for efficient electrocatalytic hydrodechlorination of recalcitrant chlorophenols and selective formation of phenol with superior Pd-mass activity (1.35 min-1 mgPd-1), which is 4.4 times of commercial Pd/C and about 10-100 times of reported Pd-based catalysts. The Pd-PdO heterostructure is stable in real water matrices and achieves selective phenol recovery (>99%) from the chlorophenol mixture and efficient detoxification along chlorophenol removal. Experimental results and computational modeling reveal that the adsorption/desorption behaviors of zerovalent Pd and PdO sites in the Pd-PdO heterostructure are optimized and a synergy is realized to promote atomic hydrogen (H*) generation, transfer, and utilization: H* is efficiently generated at zerovalent Pd sites, transferred to PdO sites, and eventually consumed in the dechlorination reaction at PdO sites. This work provides a promising strategy to realize the synergy of Pd with different valence states in the metal-metal oxide heterostructure for simultaneous decontamination, detoxification, and resource recovery from halogenated organic pollutants.

Operation regimes of spinal circuits controlling locomotion and the role of supraspinal drives and sensory feedback

Locomotion in mammals is directly controlled by the spinal neuronal network, operating under the control of supraspinal signals and somatosensory feedback that interact with each other. However, the functional architecture of the spinal locomotor network, its operation regimes, and the role of supraspinal and sensory feedback in different locomotor behaviors, including at different speeds, remain unclear. We developed a computational model of spinal locomotor circuits receiving supraspinal drives and limb sensory feedback that could reproduce multiple experimental data obtained in intact and spinal-transected cats during tied-belt and split-belt treadmill locomotion. We provide evidence that the spinal locomotor network operates in different regimes depending on locomotor speed. In an intact system, at slow speeds (<0.4 m/s), the spinal network operates in a non-oscillating state-machine regime and requires sensory feedback or external inputs for phase transitions. Removing sensory feedback related to limb extension prevents locomotor oscillations at slow speeds. With increasing speed and supraspinal drives, the spinal network switches to a flexor-driven oscillatory regime and then to a classical half-center regime. Following spinal transection, the model predicts that the spinal network can only operate in the state-machine regime. Our results suggest that the spinal network operates in different regimes for slow exploratory and fast escape locomotor behaviors, making use of different control mechanisms.

Operation regimes of spinal circuits controlling locomotion and the role of supraspinal drives and sensory feedback.

Locomotion in mammals is directly controlled by the spinal neuronal network, operating under the control of supraspinal signals and somatosensory feedback that interact with each other. However, the functional architecture of the spinal locomotor network, its operation regimes, and the role of supraspinal and sensory feedback in different locomotor behaviors, including at different speeds, remain unclear. We developed a computational model of spinal locomotor circuits receiving supraspinal drives and limb sensory feedback that could reproduce multiple experimental data obtained in intact and spinal-transected cats during tied-belt and split-belt treadmill locomotion. We provide evidence that the spinal locomotor network operates in different regimes depending on locomotor speed. In an intact system, at slow speeds (<0.4 m/s), the spinal network operates in a non-oscillating state-machine regime and requires sensory feedback or external inputs for phase transitions. Removing sensory feedback related to limb extension prevents locomotor oscillations at slow speeds. With increasing speed and supraspinal drives, the spinal network switches to a flexor-driven oscillatory regime and then to a classical half-center regime. Following spinal transection, the model predicts that the spinal network can only operate in the state-machine regime. Our results suggest that the spinal network operates in different regimes for slow exploratory and fast escape locomotor behaviors, making use of different control mechanisms.

Selective Inhibition of Deamidated Triosephosphate Isomerase by Disulfiram, Curcumin, and Sodium Dichloroacetate: Synergistic Therapeutic Strategies for T-Cell Acute Lymphoblastic Leukemia in Jurkat Cells.

T-cell acute lymphoblastic leukemia (T-ALL) is a challenging childhood cancer to treat, with limited therapeutic options and high relapse rates. This study explores deamidated triosephosphate isomerase (dTPI) as a novel therapeutic target. We hypothesized that selectively inhibiting dTPI could reduce T-ALL cell viability without affecting normal T lymphocytes. Computational modeling and recombinant enzyme assays revealed that disulfiram (DS) and curcumin (CU) selectively bind and inhibit dTPI activity without affecting the non-deamidated enzyme. At the cellular level, treatment with DS and CU significantly reduced Jurkat T-ALL cell viability and endogenous TPI enzymatic activity, with no effect on normal T lymphocytes, whereas the combination of sodium dichloroacetate (DCA) with DS or CU showed synergistic effects. Furthermore, we demonstrated that dTPI was present and accumulated only in Jurkat cells, confirming our hypothesis. Finally, flow cytometry confirmed apoptosis in Jurkat cells after treatment with DS and CU or their combination with DCA. These findings strongly suggest that targeting dTPI represents a promising and selective target for T-ALL therapy.

Effect of Hydrodynamic Shear Stress on Algal Cell Fate in 3D Extrusion Bioprinting

The 3D bioprinting of aquatic photosynthetic organisms holds potential for applications in biosensing, wastewater treatment, and biofuel production. While algae cells can be immobilized in bioprinted cell‐friendly matrices, there is a knowledge gap regarding the thresholds of hydrodynamic shear stress that affect the cells’ functionality and viability during bioprinting. This study examines the effect of hydrodynamic shear stress on the fate of Chlamydomonas reinhardtii cells. Computational fluid dynamics models based on the Navier–Stokes equations are developed to numerically predict the shear stresses experienced by the cells during extrusion. Parallelly, cell culture experiments are conducted to evaluate the functionality, growth rates, and viability of algae cells within bioprinted constructs. By correlating cell culture and simulation results, the causal link between shear stress in the nozzle and cell viability and function has been characterized. The findings highlight that cell viability and function are significantly impacted by process factors. Notably, algae cell function is more sensitive to shear stress than cell viability. Functional impairments occur at maximum shear stresses around 5 kPa, while viability remains unaffected. Beyond 14 kPa, both functionality and viability decline significantly and irreversibly. The results emphasize the importance of assessing viability and function after bioprinting, rather than just viability.

Attention mechanism‐assisted recurrent neural network for well log lithology classification

AbstractLithology classification is a fundamental aspect of reservoir classification. Due to the limited availability of core samples, computational modelling methods for lithology classification based on indirect measurements are required. The main challenge for standard clustering methods is the complex vertical dependency of sedimentological sequences as well as the spatial coupling of well logs. Machine learning methods, such as recurrent neural networks, long short‐term memory and bidirectional long short‐term memory, can account for the spatial correlation of the measured data and the predicted model. Based on these developments, we propose a novel approach using two distinct models: a self‐attention‐assisted bidirectional long short‐term memory model and a multi‐head attention‐based bidirectional long short‐term memory model. These models consider spatial continuity and adaptively adjust the weight in each step to improve the classification using the attention mechanism. The proposed method is tested on a set of real well logs with limited training data obtained from core samples. The prediction results from the proposed models and the benchmark one are compared in terms of the accuracy of lithology classification. Additionally, the weight matrices from both attention mechanisms are visualized to elucidate the correlations between depth steps and to help analyse how these mechanisms contribute to improved prediction accuracy. The study shows that the proposed multi‐head attention‐based bidirectional long short‐term memory model improves classification, especially for thin layers.

Multi-material 3D printing of functionally graded soft-hard interfaces for enhancing mandibular kinematics of temporomandibular joint replacement prostheses

Temporomandibular joint (TMJ) replacement prostheses often face limitations in accommodating translational movements, leading to unnatural kinematics and loading conditions, which affect functionality and longevity. Here, we investigate the potential of functionally graded materials (FGMs) in TMJ prostheses to enhance mandibular kinematics and reduce joint reaction forces. We develop a functionally graded artificial cartilage for the TMJ implant and evaluate five FGM designs: hard, hard-soft, and three FGM gradients with gradual transitions from 90% hard material to 0%, 10%, and 20%. These designs are 3D printed, mechanically tested under quasi-static compression, and simulated under physiological conditions. Results from computational modeling and experiments are compared to an intact mandible during incisal clenching and left group biting. The FGM design with a transition from 90% to 0% hard material improves kinematics by 19% and decreases perfomance by 3%, reduces joint reaction forces by 8% and 10%, and increases mandibular movement by 20% and 88% during incisal clenching and left group biting, respectively. These findings provide valuable insights for next-generation TMJ implants.

Cerebellar climbing fibers signal flexible, rapidly adapting reward predictions.

Classical models of cerebellar computation posit that climbing fibers (CFs) operate according to supervised learning rules, correcting movements by signaling the occurrence of motor errors. However, recent findings suggest that in some behaviors, CF activity can exhibit features that resemble the instructional signals necessary for reinforcement learning, namely reward prediction errors (rPEs). Despite these initial observations, many key properties of reward-related CF responses remain unclear, thus limiting our understanding of how they operate to guide cerebellar learning. Here, we have measured the postsynaptic responses of CFs onto cerebellar Purkinje cells using two-photon calcium imaging to test how they respond to learned stimuli that either do or do not predict reward. We find that CFs can develop generalized responses to similar cues of the same modality, regardless of whether they are reward predictive. However, this generalization depends on temporal context, and does not extend across sensory modalities. Further, learned CF responses are flexible, and can be rapidly updated according to new reward contingencies. Together these results suggest that CFs can generate learned, reward-predictive responses that flexibly adapt to the current environment in a context-sensitive manner.

Economic impacts of large dams on downstream brickmaking in developing countries

Large dams have positive and negative impacts, including disrupting brickmaking on the floodplains downstream due to flow regulation and sediment reduction, affecting the supply of essential construction material, notably in developing countries. In this study, we introduce an analytical framework to assess the economywide effects of large dams on downstream brickmaking, focusing on Traditional Fired Clay Brick (TFCB). The framework includes three steps: characterizing the impacts on river flow and sediment load using river system modeling and secondary data, understanding the role of TFCB production in the economy based on survey and economic data, and quantifying the economywide impacts of changes in TFCB production using dynamic computable general equilibrium modeling. We demonstrate the functionality of the approach by conducting a case study of the impacts of the Grand Ethiopian Renaissance Dam (GERD) on the Sudanese economy due to changes in TFCB production by comparing two scenarios: “with GERD” and “no GERD.” Results show that Sudan’s accumulated (2023–2050) discounted (at 0.5% annually) Gross Domestic Product (GDP) at factor cost would decline by US$ 6 billion (−0.38%) due to a reduction in TFCB production. Consumer flexibility regarding brick types and the ability of alternative brick sources to fill the demand gap are key determinants of the impacts.

Expectation generation and its effect on subsequent pain and visual perception.

Bayesian accounts of perception, such as predictive processing, suggest that perceptions integrate expectations and sensory experience, and thus assimilate to expected values. Furthermore, more precise expectations should have stronger influences on perception. We tested these hypotheses in a paradigm that manipulates both the mean value and the precision of cues within-person. Forty-five participants observed cues-presented as ratings from 10 previous participants-with varying cue means, variances (precision), and skewness across trials. Participants reported expectations regarding the painfulness of thermal stimuli or the visual contrast of flickering checkerboards. Subsequently, similar cues were each followed by a visual or noxious thermal stimulus. While perceptions assimilated to expected values in both modalities, cues' precision mainly affected visual ratings. Furthermore, behavioral and computational models revealed that expectations were biased towards extreme values in both modalities, and towards low-pain cues specifically. fMRI analysis revealed that the cues affected systems related to higher-level affective and cognitive processes-including assimilation to the cue mean in a neuromarker of endogenous contributions to pain and in the nucleus accumbens, and activity consistent with aversive prediction-error-like encoding in the periaqueductal gray during pain perception-but not systems related to early perceptual processing. Our findings suggest that predictive processing theories should be combined with mechanisms such as selective attention to better fit empirical findings, and that expectation generation and its perceptual effects are mostly modality-specific and operate on higher-level processes rather than early perception.

Shear Capacity Prediction of Extremely-Loaded Box Culvert on Elastic Soil Using Artificial Neural Network

A box culvert, buried at shallow depths beneath roadways, may experience deflections caused by the dynamic impact of traffic loading and the vertical pressure exerted by the soil fill. A computational model commonly employed used to various engineering issues, including those in geotechnical applications, is the beam-on-elastic-foundation model. In this context, the Moment Distribution Method (MDM) must be applied to account for the elastic foundation. To achieve this, the internal forces acting on the ends of both exterior and interior walls are transferred to the beam-like bottom slab of the culvert, which rests on an elastic soil bed. Subsequently, the secondary internal forces are determined by refining the structural parameters, taking into account the characteristics of the elastic soil bed. This study presents the development and application of an Artificial Neural Network (ANN) model to predict the shear capacity of box culverts on elastic soil under traffic loading conditions. The proposed model is trained and validated using a comprehensive database of beam on elastic foundation solutions. The input parameters include the geometrical and mechanical properties of the culvert and the soil, as well as the loading conditions. The results of the ANN model show R2 values of 0.9633 and 0.9581 for the training and testing sets, respectively, indicating the model's excellent accuracy. These findings suggest that the ANN model can reliably predict the shear capacity of culverts.

Analysis of Erosion of Surfaces in Falling Particle Concentrating Solar Power

Abstract Next generation concentrating solar power (CSP) systems, which utilize solid particles for energy capture, transport, and storage, offer prospects for higher temperature operation, improved efficiency, and reduced overall costs. Nevertheless, the continuous impingement of particles on component materials can result in substantial erosion, significantly constraining the performance and longevity of the components. A comprehensive understanding of particle erosion on surfaces is essential for designing components and operational parameters or coatings that minimize wear. This study presents a computational physics-based particle tracking model of the erosion rate of incident surfaces under different geometric, operational, and particle parameters. The computational model is validated with experimental measurements conducted as part of the study. Computational simulations are presented to elucidate the effects of each parameter and further used to investigate erosion rates in a systematic design of experiments covering a wide range of parameters. Based on the simulation results, a generalized analytical model is developed to relate erosion wear to pertinent dimensionless groups governing the physics of the process. The analytical model is shown to be accurate to within 10% and its use in understanding surface erosion as well as designing wear-resistant coatings to limit erosion within acceptable values is presented.

Progress of Conductivity and Conduction Velocity Measured in Human and Animal Hearts.

Cardiac conduction velocity (CV) is a critical electrophysiological characteristic of the myocardium, representing the speed at which electrical pulses propagate through cardiac tissue. It can be delineated into longitudinal, transverse, and normal components in the myocardium. The CV and its anisotropy ratio are crucial to both normal electrical conduction and myocardial contraction, as well as pathological conditions where it increases the risk of conduction block and reentry. This comprehensive review synthesizes longitudinal and transverse CV values from clinical and experimental studies of human infarct hearts, including findings from the isthmus and outer loop, alongside data derived from animal models. Additionally, we explore the anisotropic ratio of conductivities assessed through both animal and computational models. The review culminates with a synthesis of scientific evidence that guides the selection of CV and its corresponding conductivity in cardiac modeling, particularly emphasizing its application in patient-specific cardiac arrhythmia modeling.

An Algorithm for Creating a Synaptic Cleft Digital Phantom Suitable for Further Numerical Modeling

One of the most significant applications of mathematical numerical methods in biology is the theoretical description of the convectional reaction–diffusion of chemical compounds. Initial biological objects must be appropriately mimicked by digital domains that are suitable for further use in computational modeling. In the present study, an algorithm for the creation of a digital phantom describing a local part of nervous tissue—namely, a synaptic contact—is established. All essential elements of the synapse are determined using a set of consistent Boolean operations within the COMSOL Multiphysics software 6.1. The formalization of the algorithm involves a sequence of procedures and logical operations applied to a combination of 3D Voronoi diagrams, an experimentally defined inner synapse area, and a simple ellipsoid under different sets of biological parameters. The obtained digital phantom is universal and may be applied to different types of neuronal synapses. The clear separation of the designed domains reveals that the boundary’s conditions and internal flux dysconnectivity functions can be set up explicitly. Digital domains corresponding to the parts of a synapse are appropriate for further application of the derived numeric meshes, with various capacities of the included elements. Thus, the obtained digital phantom can be effectively used for further modeling of the convectional reaction–diffusion of chemical compounds in nervous tissue.

Thermal Conductivity Measurement System for Functional and Structural Products

An automated system for measuring the thermal conductivity of functional and structural materials was developed. The main building blocks of the setup are the following: heating unit and cooling unit creating a heat flux gradient in the test sample; thermal resistances for temperature registration and control; and thermal pads for better contact between parts of the setup and the sample. The effect of the thermal conductivity of thermal pads and thermal resistances on the distribution of thermal fields in the developed setup was studied by computer modelling. A control software for the measuring setup was developed based on the hardware implementation of the steady-state Fourier’s law-based method for the determination of thermal conductivity. The stopping criterion for the setup control software is the equality of heat fluxes in the heating and cooling units, as well as the stability of the thermal conductivity coefficient readings. The testing and calibration of the device were carried out using a sample of pure aluminum (99.999 wt.% Al). It was found that the experimental value of the thermal conductivity coefficient of the aluminum sample at room temperature (T = 22 °C) is &lt;λ&gt; = 243 ± 3 W/m·K. This value of the thermal conductivity coefficient is consistent with the literature data and experimental values obtained by the laser flash method, which ranges within λ = 210–260 W/m·K.