Candidate Models Research Articles

The missing quasar image in the gravitationally lensed quasar HE0230−2130: Implications for the cored lens mass distribution and dark satellites

Strongly lensed systems with peculiar configurations allow us to probe the local properties of the deflecting lens mass while simultaneously testing general profile assumptions. The quasar HE0230−2130 is lensed by two galaxies at similar redshifts (Δz ∼ 0.003) into four observed images. Using modeled quasar positions from fitting the brightness of the quasar images in ground-based imaging data from the Magellan telescope, we find that lens-mass models where each of these two galaxies is parametrized with a singular power-law (PL) profile predict five quasar images. One of the predicted images is unobserved despite it being distinctively offset from the lensing galaxies and likely bright enough to be observable. This missing image gives rise to new opportunities to study the mass distribution of these galaxies. To interpret the quad configuration of the system, we tested 12 different profile assumptions with the aim of obtaining lens-mass models that correctly predict only four observed images. We tested the effects of adopting: cored profiles for the lensing galaxies; external shear; and additional profiles to represent a dark matter clump. We find that half of our model classes can produce the correct image multiplicity. By comparing the Bayesian evidence of different model parametrizations, we favor two model classes: (i) one that incorporates two singular PL profiles for the lensing galaxies and a cored isothermal sphere in the region of the previously predicted fifth image (rNIS profile), and (ii) one with a bigger lensing galaxy parametrized by a singular PL profile and the smaller galaxy by a cored PL profile with external shear. We estimated the mass of the rNIS clump for each candidate model of our final Markov chain Monte Carlo sample, and find that only 2% are in the range of 106 M⊙ ≤ MrNIS ≤ 109 M⊙, which is the predicted mass range of dark matter subhalos in cold dark matter simulations, or the mass of dark-matter-dominated and low-surface-brightness galaxies. We therefore favor the models with a cored mass distribution for the lens galaxy close to the predicted fifth image. Our study further demonstrates that lensed quasar images are sensitive to the dark matter structure in the gravitational lens. We are able to describe this exotic lensing configuration with relatively simple models, which demonstrates the power of strong lensing for studying galaxies and lens substructure.

A Mallows-type model averaging estimator for de-noise linear models

. This article is concerned with model averaging for de-noise linear models, in which some covariates are not observed, but their ancillary variables are available. The least-squares-based estimation procedure is used to estimate the unknown regression parameter in each candidate model after the calibrated error-prone covariates are obtained. Then a Mallows-type weight choice criterion is constructed. When all candidate models are misspecified, the model averaging estimator is asymptotically optimal in the sense that achieving the lowest possible squared error. On the other hand, when the true model is included in the set of candidate models, the model averaging estimator of the regression parameter is root n consistent. The finite sample performance of our model averaging estimator is evaluated by some simulation studies. The proposed procedure is further applied to real-data analysis.

Water vapor content prediction based on neural network model selection and optimal fusion

Accurately predicting the evolution of water vapor content has significant practical engineering implications for accurately determining the timing of artificial rain enhancement operations. However, the nonlinear, unstable, and chaotic nature of water vapor content presents considerable challenges to its precise prediction. To address this, our research proposes a combined prediction model, based on a neural network model selection mechanism and small sample water vapor content data observed by microwave radiometers in the Zhengzhou area. This combined model utilizes seven neural network models as candidate models and dynamically selects the optimal three models as the central prediction components via evaluation indicators. Additionally, we employ the maximum information coefficient method to preserve the optimal input features and use a time-varying rate wave mode decomposition algorithm as the front-end processing module. This module processes the original water vapor content sequence and extracts relevant latent information. Finally, we incorporate the African vulture intelligent optimization algorithm as the backend processing component to optimize the weighted combination coefficients, achieving the integration of prediction results. Experimental results reveal that the Mean Absolute Percentage Errors (MAPE) of the proposed model on three types of water vapor content datasets are 0.208%, 0.876%, and 0.369%, respectively. This suggests a predictive performance improvement of at least 10% compared to existing models. The prediction model proposed in this study not only establishes an accurate and reliable mechanism for predicting water vapor content but also offers more comprehensive decision support information for weather modification operations.

Adjoint-accelerated Bayesian inference applied to the thermoacoustic behaviour of a ducted conical flame

We use Bayesian inference, accelerated by adjoint methods, to construct a quantitatively accurate model of the thermoacoustic behaviour of a conical flame in a duct. We first perform a series of automated experiments on a ducted flame rig. Next, we propose several candidate models of the rig's components and assimilate data into each model to find the most probable parameters for that model. We rank the candidate models based on their marginal likelihood (evidence) and select the most likely model for each component. We begin this process by rigorously characterizing the acoustics of the cold rig. When the flame is introduced, we propose several candidate models for the fluctuating heat release rate. We find that the most likely flame model considers velocity perturbations in both the burner feed tube and the outer duct, even though studies in the literature typically neglect either one of these. Using the most likely model, we infer the flame transfer functions for 24 flames and quantify their uncertainties. We do this with the flames in situ, using only pressure measurements. We find that the inferred flame transfer functions render the model quantitatively accurate, and, where comparable, broadly consistent with direct measurements from several studies in the literature.

Recent advances and challenges in protein complex model accuracy estimation

Estimation of model accuracy plays a crucial role in protein structure prediction, aiming to evaluate the quality of predicted protein structure models accurately and objectively. This process is not only key to screening candidate models that are close to the real structure, but also provides guidance for further optimization of protein structures. With the significant advancements made by AlphaFold2 in monomer structure, the problem of single-domain protein structure prediction has been widely solved. Correspondingly, the importance of assessing the quality of single-domain protein models decreased, and the research focus has shifted to estimation of model accuracy of protein complexes. In this review, our goal is to provide a comprehensive overview of the reference and statistical metrics, as well as representative methods, and the current challenges within four distinct facets (Topology Global Score, Interface Total Score, Interface Residue-Wise Score, and Tertiary Residue-Wise Score) in the field of complex EMA.

Data-Driven Identification of Crane Dynamics Using Regularized Genetic Programming

The meaningful problem of improving crane safety, reliability, and efficiency is extensively studied in the literature and targeted via various model-based control approaches. In recent years, crane data-driven modeling has attracted much attention compared to physics-based models, particularly due to its potential in real-time crane control applications, specifically in model predictive control. This paper proposes grammar-guided genetic programming with sparse regression (G3P-SR) to identify the nonlinear dynamics of an underactuated crane system. G3P-SR uses grammars to bias the search space and produces a fixed number of candidate model terms, while a local search method based on an l0-regularized regression results in a sparse solution, thereby also reducing model complexity as well as reducing the probability of overfitting. Identification is performed on experimental data obtained from a laboratory-scale overhead crane. The proposed method is compared with multi-gene genetic programming (MGGP), NARX neural network, and Takagi-Sugeno fuzzy (TSF) ARX models in terms of model complexity, prediction accuracy, and sensitivity. The G3P-SR algorithm evolved a model with a maximum mean square error (MSE) of crane velocity and sway prediction of 1.1860 × 10−4 and 4.8531 × 10−4, respectively, in simulations for different testing data sets, showing better accuracy than the TSF ARX and MGGP models. Only the NARX neural network model with velocity and sway maximum MSEs of 1.4595 × 10−4 and 4.8571 × 10−4 achieves a similar accuracy or an even better one in some testing scenarios, but at the cost of increasing the total number of parameters to be estimated by over 300% and the number of output lags compared to the G3P-SR model. Moreover, the G3P-SR model is proven to be notably less sensitive, exhibiting the least deviation from the nominal trajectory for deviations in the payload mass by approximately a factor of 10.

Statistical considerations in model-based dose finding for binary responses under model uncertainty.

The statistical methodology for model-based dose finding under model uncertainty has attracted increasing attention in recent years. While the underlying principles are simple and easy to understand, developing and implementing an efficient approach for binary responses can be a formidable task in practice. Motivated by the statistical challenges encountered in a phase II dose finding study, we explore several key design and analysis issues related to the hybrid testing-modeling approaches for binary responses. The issues include candidate model selection and specifications, optimal design and efficient sample size allocations, and, notably, the methods for dose-response testing and estimation. Specifically, we consider a class of generalized linear models suited for the candidate set and establish D-optimal designs for these models. Additionally, we propose using permutation-based tests for dose-response testing to avoid asymptotic normality assumptions typically required for contrast-based tests. We perform trial simulations to enhance our understanding of these issues.

Exploring the chemical space of BRAF Inhibitors: A cheminformatic and Machine learning analysis

BRAF kinase inhibitors have shown promise in treating melanoma patients with BRAF-V600 mutations. However, their clinical benefits are limited due to the development of acquired resistance. To address this issue, we conducted a systematic cheminformatic analysis and machine learning modeling to study the chemical space, scaffolds, structure–activity relationship, and landscape of human BRAF inhibitors. The final dataset comprised 3,952 molecules. Physicochemical property visualization for chemical space visualization has demonstrated that molecules from Group 1 (potent/active class) generally have slightly higher MW, RB, NumHAcceptors, and TPSA than molecules from Group 2 (intermediate/inactive class). The principal component analysis (PCA) shows that Group 1 data spreads widely and overlaps with the other group, which occupies only the left area on the plot. This suggests that molecules’ chemical structure or scaffolds are more diverse for Group 1 (Potent and Active) compounds than for Group 2 (Intermediate and Inactive) compounds. Murcko scaffold analysis has shown a greater scaffold diversity in the Active, Intermediate, and Inactive classes than in the Potent class. However, in the four bioactivity-defined classes for our BRAF dataset, molecules are uniformly distributed, with a small number of highly populated scaffolds and a large number of singletons. Furthermore, scaffold visualization has identified 12 representative Murcko scaffolds. Scaffolds 1 S1C1, S2C1, S2C2, S3C2, and S4C2 are highly favorable due to their high scaffold enrichment factor values. Based on scaffold analysis, the study investigated and summarized the local structure–activity relationships (SARs). In addition, the global SAR landscape was explored through quantitative structure–activity relationship (QSAR) modeling and structure–activity landscape visualization. Out of a total of 14 candidate models for BRAF inhibitors, a QSAR classification model that includes all 3952 molecules has been identified as the best model. The model was built using the PubChem fingerprint and extra trees algorithm and achieved an accuracy of 0.920 for the training set, 0.699 for the 10-fold cross-validation set, and 0.733 for the test set. Through a detailed analysis of the structure–activity landscapes, a total of sixteen significant consensus activity cliff (AC) generators were identified (ChEMBL molecule IDs: 4795335, 1822247, 3665863, 5083036, 5086749, 4061139, 1822242, 3665859, 1822250, 3641129, 3641043, 514688,3641178, 3697934, 3641046, and 368991). These generators offer valuable information on the SAR (Structure-Activity Relationship) for medicinal chemistry. The findings from this study provide new insights and guidelines for hit identification and lead optimization in developing novel BRAF inhibitors.

Spatial-temporal Bayesian accelerated failure time models for survival endpoints with applications to prostate cancer registry data

Prostate cancer is the most common cancer after non-melanoma skin cancer and the second leading cause of cancer deaths in US men. Its incidence and mortality rates vary substantially across geographical regions and over time, with large disparities by race, geographic regions (i.e., Appalachia), among others. The widely used Cox proportional hazards model is usually not applicable in such scenarios owing to the violation of the proportional hazards assumption. In this paper, we fit Bayesian accelerated failure time models for the analysis of prostate cancer survival and take dependent spatial structures and temporal information into account by incorporating random effects with multivariate conditional autoregressive priors. In particular, we relax the proportional hazards assumption, consider flexible frailty structures in space and time, and also explore strategies for handling the temporal variable. The parameter estimation and inference are based on a Monte Carlo Markov chain technique under a Bayesian framework. The deviance information criterion is used to check goodness of fit and to select the best candidate model. Extensive simulations are performed to examine and compare the performances of models in different contexts. Finally, we illustrate our approach by using the 2004-2014 Pennsylvania Prostate Cancer Registry data to explore spatial-temporal heterogeneity in overall survival and identify significant risk factors.

Cascade & allocate: A cross-structure adversarial attack against models fusing vision and language

The data fusion systems between multiple modals have attracted a wide range of concerns about their adversarial robustness. Recent image captioning attacks lack cross-structure transferability against models with diverse structures. Therefore, the attackers demonstrate satisfactory performance only when they gain full or partial knowledge of target image captioning models. So far, the cross-structure transferability of multi-modal adversarial examples has not been thoroughly investigated. In this paper, a theorem is proposed to analyze the upper bound captioning adversarial transferability. We design a new transfer-based adversarial attack method (Cascade & Allocate) against image captioning models. Our method can be divided into two steps. The first step randomly selects a set of candidate models with diverse structures, and we use a momentum-like strategy to generate perturbations. This ‘model cascade’ step narrows the gap of the gradient directions between different image captioning models, therefore enhancing the cross-model transferability and generating model-agnostic adversarial perturbations. The second step utilizes the upper bound transferability theorem to allocate the weight of perturbations per model. This ‘weight allocate’ step enhances the weight of model with the most consistent gradient in all candidate models, which generates more transferable adversarial examples. In the experiments, we compare our approach with other captioning and ensemble-based attacks against five black-box models with different structures on MS COCO and Flickr-30k datasets. The adversarial examples exhibit state-of-the-art adversarial transferability against five black-box models with different structures. The results demonstrate that our approach is practical and generalizes well to a wide range of captioning models.

Personalizing patient risk of a life-altering event: An application of machine learning to hemiarch surgery

ObjectiveThe study objective was to assess a machine learning model’s ability to predict the occurrence of life-altering events in hemiarch surgery and determine contributing patient characteristics and intraoperative factors. MethodsIn total, 602 patients who underwent hemiarch replacement at a high-volume aortic center from 2009 to 2022 were included. Patients were randomly divided into training (80%) and testing (20%) sets with various eXtreme gradient boosting candidate models constructed to predict the risk of experiencing life-altering events, including stroke, mortality, or new renal replacement therapy requirement. A total of 64 input parameters from the index hospitalization were identified, including 24 demographic characteristics as well as 8 preoperative and 32 intraoperative variables. A SHapley Additive exPlanation beeswarm plot was generated to identify and interpret the impact of individual features on the predictions of the final model. ResultsA life-altering event was noted in 15% (90/602) of patients who underwent hemiarch replacement, including urgent/emergency cases and dissections. The final eXtreme Gradient Boosting model demonstrated a cross-validation accuracy of 88% on the testing set and was well calibrated as evidenced by a low Brier score of 0.12. The best performing model achieved an area under the receiver operating characteristic curve of 0.76 and an area under the precision recall curve of 0.55. The SHapley Additive exPlanation beeswarm plot provided insights into key features that significantly influenced model prediction. ConclusionsMachine learning demonstrated superior accuracy in predicting hemiarch patients who would experience a life-altering event. This model may help to guide patients and clinicians in stratifying risk on an individual basis, which may in turn influence clinical decision-making.

The local to unity dynamic Tobit model

This paper considers highly persistent time series that are subject to nonlinearities in the form of censoring or an occasionally binding constraint, such as are regularly encountered in macroeconomics. A tractable candidate model for such series is the dynamic Tobit with a root local to unity. We show that this model generates a process that converges weakly to a non-standard limiting process, that is constrained (regulated) to be positive. Surprisingly, despite the presence of censoring, the OLS estimators of the model parameters are consistent. We show that this allows OLS-based inferences to be drawn on the overall persistence of the process (as measured by the sum of the autoregressive coefficients), and for the null of a unit root to be tested in the presence of censoring. Our simulations illustrate that the conventional ADF test substantially over-rejects when the data is generated by a dynamic Tobit with a unit root, whereas our proposed test is correctly sized. We provide an application of our methods to testing for a unit root in the Swiss franc/euro exchange rate, during a period when this was subject to an occasionally binding lower bound.

Ontogeny of the autopodial skeleton of the gecko Tarentola (Squamata: Phyllodactylidae)

Squamates exhibit evident diversity in their limb morphology. Gekkotans are a particularly diverse group in this respect. The appearance of toepads in gekkotans usually cooccurs with the reduction or loss of claws. The gecko Tarentola (Phyllodactylidae) shows a unique combination of features among geckos, with toepads, hyperphalangy, and dimorphism of claw expression (claws are retained on digits III and IV, but lost (manus) or strongly reduced (pes) on the remaining digits). Despite being a candidate model for studying embryonic skeletal development of the autopodium, no studies have investigated the autopodial development of the gecko Tarentola in detail. Here, we aim to follow up the development of the autopodial skeleton in T. annularis and T. mauritanica using acid-free double staining. The results indicate that the terminal phalanges of claw-bearing digits III and IV ossify earlier than in the remaining digits. This confirms the differential ossification as a result of claw regression in Tarentola. The strongly reduced second phalanges of digits IV in both the fore- and hindlimbs are the last ossifying phalanges. Such late ossification may precede the evolutionary loss of this phalanx. If this is correct, the autopodia of Tarentola would be an interesting example of both the hyperphalangy in digit I and the process of phalanx loss in digit IV. Delay in ossification of the miniaturised phalanx probably represents an example of paedomorphosis.

CNN-based model updating for structures by direct use of dynamic structural response measurements

This study presents a convolutional neural network (CNN)-based model updating method for structures using dynamic structural responses. In the presented method, the dynamic structural response is used directly for model updating, thereby omitting the manual process of extracting modal parameters. In the method, the type of stiffness variable in the target structure for model updating is selected, and a number of candidate models are generated by multiple selected stiffness variables in the structures and possible values of variables. Dynamic structural responses are then extracted by time history structural analysis of the candidate models, and a CNN sets dynamic structural response as input, and information on stiffness variables as output is introduced to learn the relationship between the two data. CNNs trained using dynamic displacement response and dynamic acceleration response are presented as Model_XT and Model_XTa, respectively. The presented models can rapidly estimate stiffness of the target structure by entering dynamic responses into the models. These models are applied to model updating for three example structures, and the stiffness estimation performance for model updating is evaluated. Furthermore, the length of data used for CNN training, the stiffness estimation performance according to the CNN input layer settings, and the stiffness estimation applied with limited data to the updated model are analyzed.

Resisting Backdoor Attacks in Federated Learning via Bidirectional Elections and Individual Perspective

Existing approaches defend against backdoor attacks in federated learning (FL) mainly through a) mitigating the impact of infected models, or b) excluding infected models. The former negatively impacts model accuracy, while the latter usually relies on globally clear boundaries between benign and infected model updates. However, in reality, model updates can easily become mixed and scattered throughout due to the diverse distributions of local data. This work focuses on excluding infected models in FL. Unlike previous perspectives from a global view, we propose Snowball, a novel anti-backdoor FL framework through bidirectional elections from an individual perspective inspired by one principle deduced by us and two principles in FL and deep learning. It is characterized by a) bottom-up election, where each candidate model update votes to several peer ones such that a few model updates are elected as selectees for aggregation; and b) top-down election, where selectees progressively enlarge themselves through picking up from the candidates. We compare Snowball with state-of-the-art defenses to backdoor attacks in FL on five real-world datasets, demonstrating its superior resistance to backdoor attacks and slight impact on the accuracy of the global model.

Dynamic Knowledge Injection for AIXI Agents

Prior approximations of AIXI, a Bayesian optimality notion for general reinforcement learning, can only approximate AIXI's Bayesian environment model using an a-priori defined set of models. This is a fundamental source of epistemic uncertainty for the agent in settings where the existence of systematic bias in the predefined model class cannot be resolved by simply collecting more data from the environment. We address this issue in the context of Human-AI teaming by considering a setup where additional knowledge for the agent in the form of new candidate models arrives from a human operator in an online fashion. We introduce a new agent called DynamicHedgeAIXI that maintains an exact Bayesian mixture over dynamically changing sets of models via a time-adaptive prior constructed from a variant of the Hedge algorithm. The DynamicHedgeAIXI agent is the richest direct approximation of AIXI known to date and comes with good performance guarantees. Experimental results on epidemic control on contact networks validates the agent's practical utility.

Research on cooling load estimation through optimal hybrid models based on Naive Bayes

Cooling load estimation is crucial for energy conservation in cooling systems, with applications like advanced air-conditioning control and chiller optimization. Traditional methods include energy simulation and regression analysis, but artificial intelligence outperforms them. Artificial intelligence models autonomously capture complex patterns, adapt, and scale with more data. They excel at predicting cooling loads influenced by various factors, like weather, building materials, and occupancy, leading to dynamic, responsive predictions and energy optimization. Traditional methods simplify real-world complexities, highlighting artificial intelligence’s role in precise cooling load forecasting for energy-efficient building management. This study evaluates Naive Bayes-based models for estimating building cooling load consumption. These models encompass a single model, one optimized with the Mountain Gazelle Optimizer and another optimized with the horse herd optimization algorithm. The training dataset consists of 70% of the data, which incorporates eight input variables related to the geometric and glazing characteristics of the buildings. Following the validation of 15% of the dataset, the performance of the remaining 15% is tested. Based on analysis through evaluation metrics, among the three candidate models, Naive Bayes optimized with the Mountain Gazelle Optimizer (NBMG) demonstrates remarkable accuracy and stability, reducing prediction errors by an average of 18% and 31% compared to the other two models (NB and NBHH) and achieving a maximum R2 value of 0.983 for cooling load prediction.

Mathematical Modeling of Dengue Cases in Region XI Philippines

Dengue was recognized by the World Health Organization (WHO) in 2023 as a prominent arthropod-borne viral illness and poses a significant global health threat. This study focuses on Region XI in the Philippines, where dengue is alarmingly prevalent, and in forecasting dengue cases, it employs the ARIMA model. The study, spanning from January 2019 to August 2023, integrates univariate exploratory data analysis and statistical methods to understand the patterns and trends of dengue cases. The ARIMA model, specifically (0,0,1) with a nonzero mean, the candidate model, was identified and utilized from RStudio through coefficient tests of AIC values. The forecasting results predict a decline in dengue cases over the next 16 months. The normality test of residuals and Q-Q plot affirm the model’s reliability. A significant test using the two-sample t-test demonstrates a substantial difference between actual and forecasted values. In conclusion, this study provides crucial insights for public health planning, community intervention, and future research. The ARIMA model’s successful application emphasizes the need for refined dengue control and mitigation strategies in Region XI. The results underscore the urgency of collective efforts to minimize dengue transmission and address the challenges posed by this prevalent and impactful disease.

Life-history traits of an exploited skate: The short-tail yellownose skate Zearaja brevicaudata (Chondrichthyes: Rajiformes).

The short-tail yellownose skate, Zearaja brevicaudata, occurs off southern Brazil and along the Argentine continental shelf, including waters surrounding the Islas Malvinas and the upper slope, and is very frequently caught by trawl fisheries throughout its range. The age, growth, and maturity of Z. brevicaudata were investigated using 204 individuals collected by commercial fishing vessels from northern Argentina, between March 2016 and September 2019. Age was determined on a sample comprising vertebrae from 151 individuals, 60 males, ranging in size from 391 to 956 mm total length (TL), and 91 females, ranging in size from 324 to 1060 mm TL. Maximum ages determined for males and females were 22 and 29 years, respectively. A Bayesian framework was employed with a set of three candidate models to estimate growth parameters (von Bertalanffy, Gompertz, and Logistic). The von Bertalanffy model had the best fit, and the sexes showed undistinguishable growth parameters. For sexes combined, derived growth mean parameters (± S.D.) were L∞ = 1081 mm ± 64.34 mm, k = 0.09 ± 0.01 years-1, and L0 = 248 mm ± 23.52 mm. The age at maturity was estimated at 13.15 and 14.66 years for males and females, respectively.

A Proof-of-Concept for a Hypolipidemic Brown Trout Model.

The impacts of hypolipidemic pharmaceuticals on fish lipid metabolism remain unexplored. However, data points to similar effects and mechanisms of action between fish and humans. Therefore, fish may be a strong model for screening hypolipidemic drug candidates and water pollution by lipid-modulating agents. This study aimed to test a new hypolipidemic model assay with juvenile brown trout using atorvastatin (ATV)-a hypolipidemic chemical. We selected 17α-ethinylestradiol (EE2), known to cause hyperlipidemia in fish, to ensure model functionality. Fish received intramuscular injections of 4 μL/g for two weeks under the following experimental conditions: control-C (0.7% NaCl), solvent control-SC (0.7% NaCl, 0.9% ethanol, 0.1% dimethyl sulfoxide), ATV (0.3 μg/g), EE2 (2 μg/g), and a mixture of both compounds-MIX (0.3 μg/g ATV and 2 μg/g EE2). Endpoints included blood lipid biochemistry, hepatic lipid droplet quantification, and liver mRNA expression of lipid-related target genes (related to lipogenesis, lipid transport, and β-oxidation pathways). ATV lowered blood total cholesterol, high-density lipoproteins (HDL), and low-density lipoproteins (LDL) levels, whilst triglycerides and very-low-density lipoproteins (VLDL) were highest under EE2. Hepatic lipid droplet deposition significantly increased in the ATV, EE2, and MIX groups. ATV and MIX caused a significant downregulation of the peroxisome proliferator-activated receptor γ (pparγ) and acetyl Co-A oxidase 3 (acox3). EE2 upregulated acyl-CoA long-chain synthetase 1 (acsl1) and downregulated both fatty acid binding protein 1 (fabp1) and acetyl Co-A oxidase 1-3I (acox1-3I). ATV caused hypolipidemic effects in juvenile brown trout and could even counteract EE2-stimulated hyperlipidemia, reinforcing the potential of fish hypo- and hyperlipidemic models.