Dimensionality Reduction Model Research Articles

Machine Learning Techniques and Hybrid Feature Selection Methods for Efficient Prediction of Cancer

The high-dimensional genomic data presents significant challenges, and traditional analytical methods often struggle to capture the complex, non-linear relationships within these datasets. This study elaborates into the application of machine learning methods for dimensionality reduction and predictive modeling of binary phenotypes using gene expression data. Various dimensionality reduction techniques are explored, including t-distributed stochastic neighbor embedding (t-SNE), Non-negative matrix factorization (NMF), Principal component analysis (PCA), and manifold learning methods. Additionally, various algorithms such as logistic regression, random forests, support vector machines (SVMs)cand naive Bayes models are evaluated for predicting phenotypes. The study employs rigorous cross-validation, permutation testing, and evaluation metrics like the Matthews Correlation Coefficient (MCC) to assess model performance. The study rigorously assesses current genomics strategies, pinpointing their drawbacks and suggesting areas for future investigation, while delving into the potential of machine learning to overcome these hurdles and offer valuable insights in genomics. Keywords- Gene Expression , Dimensionality Reduction, Biomarkers, Cancer Prediction, Epigenetics, Machine Learning, Feature extraction.

Interaction Between Hypoxia-Inducible Factor 1-alpha Gene Polymorphism and Helicobacter pylori Infection on Gastric Cancer in a Chinese Tibetan Population.

To investigate the impact of four single nucleotide polymorphisms (SNPs) of the HIF1α gene and its interaction with Helicobacter pylori (H. pylori) infection on susceptibility to gastric cancer (GC).Logistic regression was used to test the relationship between four SNPs of HIF1α gene and the susceptibility of GC. A generalized multifactor dimensionality reduction (GMDR) model was used to assess the HIF1α gene-H. pylori infection interaction.Logistic regression analysis indicated that both the rs11549465-CT genotype and the T allele were associated with an increased risk of GC, adjusted OR (95% CI) were 1.63 (1.09-2.20) (CT vs. CC) and 1.70 (1.13-2.36) (T vs. C), respectively. We also found that both the rs11549467-A allele and rs11549467-GA genotype were associated with an increased risk of GC, and adjusted OR (95% CI) were 2.21 (1.61-2.86) (GA vs. GG), 2.13 (1.65-2.65) (A vs. G), respectively. However, no statistically significant impact of rs2057482 or rs1957757 on risk of GC was found. The GMDR model indicated a statistically significant two-dimensional model combination (including rs11549467 and H. pylori infection). The selected model had testing balanced accuracy of 0.60 and the best cross-validation consistencies of 10/10 (p = 0.0107). Compared with H. pylori infection negative participants with rs11549467-GG genotype, H. pylori positive participants with the rs11549467-GA genotype had the highest GC risk, the OR (95% CI) was 3.04 (1.98-4.12).The rs11549467-A allele and rs11549467-GA genotype was associated with increased GC risk. Additionally, the gene-environment interaction between HIF-1α-rs11549467 and H. pylori infection was also correlated with an increased risk of GC.

Preoperative CT-based radiomic prognostic index to predict the benefit of postoperative radiotherapy in patients with non-small cell lung cancer: a multicenter study

BackgroundThe value of postoperative radiotherapy (PORT) for patients with non-small cell lung cancer (NSCLC) remains controversial. A subset of patients may benefit from PORT. We aimed to identify patients with NSCLC who could benefit from PORT.MethodsPatients from cohorts 1 and 2 with pathological Tany N2 M0 NSCLC were included, as well as patients with non-metastatic NSCLC from cohorts 3 to 6. The radiomic prognostic index (RPI) was developed using radiomic texture features extracted from the primary lung nodule in preoperative chest CT scans in cohort 1 and validated in other cohorts. We employed a least absolute shrinkage and selection operator-Cox regularisation model for data dimension reduction, feature selection, and the construction of the RPI. We created a lymph-radiomic prognostic index (LRPI) by combining RPI and positive lymph node number (PLN). We compared the outcomes of patients who received PORT against those who did not in the subgroups determined by the LRPI.ResultsIn total, 228, 1003, 144, 422, 19, and 21 patients were eligible in cohorts 1–6. RPI predicted overall survival (OS) in all six cohorts: cohort 1 (HR = 2.31, 95% CI: 1.18–4.52), cohort 2 (HR = 1.64, 95% CI: 1.26–2.14), cohort 3 (HR = 2.53, 95% CI: 1.45–4.3), cohort 4 (HR = 1.24, 95% CI: 1.01–1.52), cohort 5 (HR = 2.56, 95% CI: 0.73–9.02), cohort 6 (HR = 2.30, 95% CI: 0.53–10.03). LRPI predicted OS (C-index: 0.68, 95% CI: 0.60–0.75) better than the pT stage (C-index: 0.57, 95% CI: 0.50–0.63), pT + PLN (C-index: 0.58, 95% CI: 0.46–0.70), and RPI (C-index: 0.65, 95% CI: 0.54–0.75). The LRPI was used to categorize individuals into three risk groups; patients in the moderate-risk group benefited from PORT (HR = 0.60, 95% CI: 0.40–0.91; p = 0.02), while patients in the low-risk and high-risk groups did not.ConclusionsWe developed preoperative CT-based radiomic and lymph-radiomic prognostic indexes capable of predicting OS and the benefits of PORT for patients with NSCLC.

Changes in Lipid Profiles with the Progression of Pregnancy in Black Women.

Background/Objectives: Lipid metabolism plays an important role in maternal health and fetal development. There is a gap in the knowledge of how lipid metabolism changes during pregnancy for Black women who are at a higher risk of adverse outcomes. We hypothesized that the comprehensive lipidome profiles would show variation across pregnancy indicative of requirements during gestation and fetal development. Methods: Black women were recruited at prenatal clinics. Plasma samples were collected at 8-18 weeks (T1), 22-29 weeks (T2), and 30-36 weeks (T3) of pregnancy. Samples from 64 women who had term births (≥37 weeks gestation) were subjected to "shotgun" Orbitrap mass spectrometry. Mixed-effects models were used to quantify systematic changes and dimensionality reduction models were used to visualize patterns and identify reliable lipid signatures. Results: Total lipids and major lipid classes showed significant increases with the progression of pregnancy. Phospholipids and glycerolipids exhibited a gradual increase from T1 to T2 to T3, while sphingolipids and total sterol lipids displayed a more pronounced increase from T2 to T3. Acylcarnitines, hydroxy acylcarnitines, and Lyso phospholipid levels significantly decreased from T1 to T3. A deviation was that non-esterified fatty acids decreased from T1 to T2 and increased again from T2 to T3, suggestive of a potential role for these lipids during the later stages of pregnancy. The fatty acids showing this trend included key fatty acids-non-esterified Linoleic acid, Arachidonic acid, Alpha-linolenic acid, Eicosapentaenoic acid, Docosapentaenoic acid, and Docosahexaenoic acid. Conclusions: Mapping lipid patterns and identifying lipid signatures would help develop intervention strategies to reduce perinatal health disparities among pregnant Black women.

A Graph-Based Mathematical Model for More Efficient Dimensionality Reduction of Landmark Data in Geometric Morphometrics

A Graph-Based Mathematical Model for More Efficient Dimensionality Reduction of Landmark Data in Geometric Morphometrics

A deep learning-based workflow for fast prediction of 3D state variables in geological carbon storage: A dimension reduction approach

Deep learning (DL) models are extensively used as surrogate models for high-fidelity simulations of multiphase fluid flow in porous media at large scales, enabling fast forecasts of the spatial–temporal evolution of three-dimensional (3D) state variables in geological carbon storage (GCS). However, training these models in high-dimensional space remains computationally demanding and prone to overfitting because of limited training data. This paper presents a novel workflow to address these challenges by integrating dimension reduction (DR) methods. The proposed workflow employed pre-trained DR models to extract the latent variables of geological models and state variables and utilized the multi-layer perceptron (MLP) for constructing mapping functions between the input and output variables in latent spaces. Subsequently, the pre-trained reconstruction models converted the MLP-predicted latent state variables to their original high-dimensional form. Furthermore, we proposed a novel strategy for the DR and reconstruction of 3D saturation fields to account for the unique data characteristics of sparsity, nonuniformity, and discontinuity. The proposed strategy applied PCA and inverse PCA for 2D average saturation fields and developed a DL-based 3D reconstruction model, leveraging three 2D average saturation fields as input to produce a 3D saturation field as output. The pre-training of DR and reconstruction models and training of MLP models were conducted on 84 Gulf of Mexico (GoM) simulations and evaluated on 12 testing simulations. Each simulation contained 720 monthly time steps, with the first 360 months as the injection period and the rest as the post-injection period. The proposed workflow, incorporating DR and DL models, accurately predicts the normalized 3D pressure fields, achieving mean square error (MSE) of 2.92 × 10−7 compared to the ground truth obtained from a full-physics simulator. Furthermore, the proposed strategy outperformed PCA and convolutional autoencoder (CAE) models on 3D saturation fields, resulting in minor workflow prediction errors with an MSE of 2.93 × 10−5. The results suggest the proposed workflow provides sufficient predictive fidelity across temporal and spatial scales, and enables a speedup of 160 times compared to the full-physics simulator, facilitating improved decision-making and risk assessment for large-scale GCS management in real-time scenarios.

ScCRT: a contrastive-based dimensionality reduction model for scRNA-seq trajectory inference.

Trajectory inference is a crucial task in single-cell RNA-sequencing downstream analysis, which can reveal the dynamic processes of biological development, including cell differentiation. Dimensionality reduction is an important step in the trajectory inference process. However, most existing trajectory methods rely on cell features derived from traditional dimensionality reduction methods, such as principal component analysis and uniform manifold approximation and projection. These methods are not specifically designed for trajectory inference and fail to fully leverage prior information from upstream analysis, limiting their performance. Here, we introduce scCRT, a novel dimensionality reduction model for trajectory inference. In order to utilize prior information to learn accurate cells representation, scCRT integrates two feature learning components: a cell-level pairwise module and a cluster-level contrastive module. The cell-level module focuses on learning accurate cell representations in a reduced-dimensionality space while maintaining the cell-cell positional relationships in the original space. The cluster-level contrastive module uses prior cell state information to aggregate similar cells, preventing excessive dispersion in the low-dimensional space. Experimental findings from 54 real and 81 synthetic datasets, totaling 135 datasets, highlighted the superior performance of scCRT compared with commonly used trajectory inference methods. Additionally, an ablation study revealed that both cell-level and cluster-level modules enhance the model's ability to learn accurate cell features, facilitating cell lineage inference. The source code of scCRT is available at https://github.com/yuchen21-web/scCRT-for-scRNA-seq.

Enhancement of Classifier Performance with Adam and RanAdam Hyper-Parameter Tuning for Lung Cancer Detection from Microarray Data-In Pursuit of Precision.

Microarray gene expression analysis is a powerful technique used in cancer classification and research to identify and understand gene expression patterns that can differentiate between different cancer types, subtypes, and stages. However, microarray databases are highly redundant, inherently nonlinear, and noisy. Therefore, extracting meaningful information from such a huge database is a challenging one. The paper adopts the Fast Fourier Transform (FFT) and Mixture Model (MM) for dimensionality reduction and utilises the Dragonfly optimisation algorithm as the feature selection technique. The classifiers employed in this research are Nonlinear Regression, Naïve Bayes, Decision Tree, Random Forest and SVM (RBF). The classifiers' performances are analysed with and without feature selection methods. Finally, Adaptive Moment Estimation (Adam) and Random Adaptive Moment Estimation (RanAdam) hyper-parameter tuning techniques are used as improvisation techniques for classifiers. The SVM (RBF) classifier with the Fast Fourier Transform Dimensionality Reduction method and Dragonfly feature selection achieved the highest accuracy of 98.343% with RanAdam hyper-parameter tuning compared to other classifiers.

Gene‒environment interaction effect of hypothalamic‒pituitary‒adrenal axis gene polymorphisms and job stress on the risk of sleep disturbances.

Studies have shown that chronic exposure to job stress may increase the risk of sleep disturbances and that hypothalamic‒pituitary‒adrenal (HPA) axis gene polymorphisms may play an important role in the psychopathologic mechanisms of sleep disturbances. However, the interactions among job stress, gene polymorphisms and sleep disturbances have not been examined from the perspective of the HPA axis. This study aimed to know whether job stress is a risk factor for sleep disturbances and to further explore the effect of the HPA axis gene × job stress interaction on sleep disturbances among railway workers. In this cross-sectional study, 671 participants (363 males and 308 females) from the China Railway Fuzhou Branch were included. Sleep disturbances were evaluated with the Pittsburgh Sleep Quality Index (PSQI), and job stress was measured with the Effort-Reward Imbalance scale (ERI). Generalized multivariate dimensionality reduction (GMDR) models were used to assess gene‒environment interactions. We found a significant positive correlation between job stress and sleep disturbances (P < 0.01). The FKBP5 rs1360780-T and rs4713916-A alleles and the CRHR1 rs110402-G allele were associated with increased sleep disturbance risk, with adjusted ORs (95% CIs) of 1.75 [1.38-2.22], 1.68 [1.30-2.18] and 1.43 [1.09-1.87], respectively. However, the FKBP5 rs9470080-T allele was a protective factor against sleep disturbances, with an OR (95% CI) of 0.65 [0.51-0.83]. GMDR analysis indicated that under job stress, individuals with the FKBP5 rs1368780-CT, rs4713916-GG, and rs9470080-CT genotypes and the CRHR1 rs110402-AA genotype had the greatest risk of sleep disturbances. Individuals carrying risk alleles who experience job stress may be at increased risk of sleep disturbances. These findings may provide new insights into stress-related sleep disturbances in occupational populations.

Gestational testosterone excess early to mid-pregnancy disrupts maternal lipid homeostasis and activates biosynthesis of phosphoinositides and phosphatidylethanolamines in sheep

Gestational hyperandrogenism is a risk factor for adverse maternal and offspring outcomes with effects likely mediated in part via disruptions in maternal lipid homeostasis. Using a translationally relevant sheep model of gestational testosterone (T) excess that manifests maternal hyperinsulinemia, intrauterine growth restriction (IUGR), and adverse offspring cardiometabolic outcomes, we tested if gestational T excess disrupts maternal lipidome. Dimensionality reduction models following shotgun lipidomics of gestational day 127.1 ± 5.3 (term 147 days) plasma revealed clear differences between control and T-treated sheep. Lipid signatures of gestational T-treated sheep included higher phosphoinositides (PI 36:2, 39:4) and lower acylcarnitines (CAR 16:0, 18:0, 18:1), phosphatidylcholines (PC 38:4, 40:5) and fatty acids (linoleic, arachidonic, Oleic). Gestational T excess activated phosphatidylethanolamines (PE) and PI biosynthesis. The reduction in key fatty acids may underlie IUGR and activated PI for the maternal hyperinsulinemia evidenced in this model. Maternal circulatory lipids contributing to adverse cardiometabolic outcomes are modifiable by dietary interventions.

A Setting Optimization Ensemble for a Distributed Power Grid Protective Relay

To ensure a stable and reliable power supply, the valid and timely response of protective relays are indispensable. Through the prevention of fault expansions, potential equipment damage or system collapse can be averted, where their setting is one vital prerequisite for such effective implementations. However, the increasing complexity of distribution power systems results in more challenges for protection tuning strategies. Ergo, this paper presents an ensemble that combines the independent factor evaluation (IFE) and quantum genetic optimization (QGO) models to further optimize the performance of relays according to their distributed tuning environment. In this ensemble, both near and far-end fault characteristics can be incorporated. In the first stage, the IFE dimensional reduction model is deployed for massive heterogeneous input data, where the statistical independence of input signals is calculated, the linear transformation matrix to decouple mixed signals is found, the linear combination of such signals is formed, and the non-Gaussian property to sort them is established. This can ameliorate the following calculation efficiency under those high-dimensional data scenarios. Subsequently, the QGO model is designed to further improve relay settings, where qubit representation is built to reduce required chromosomes, the linear superposition of the optimal solution probability in different states is implemented for a better diversity and convergence performance, and a self-adaption quantum gate is established to dynamically update the qubit chromosome groups and two-state solution combinations. Lastly, an empirical case study is presented, which validates the enhanced convergence, accuracy, and rapidity of the proposed ensemble.

Machine Learning Deciphered Molecular Mechanistics with Accurate Kinetic and Thermodynamic Prediction.

Time-lagged independent component analysis (tICA) and the Markov state model (MSM) have been extensively employed for extracting conformational dynamics and kinetic community networks from unbiased trajectory ensembles. However, these techniques may not be the optimal choice for elucidating transition mechanisms within low-dimensional representations, especially for intricate biosystems. Unraveling the association mechanism in such complex systems always necessitates permutations of several essential independent components or collective variables, a process that is inherently obscure and may require empirical knowledge for selection. To address these challenges, we have implemented an integrated unsupervised dimension reduction model: uniform manifold approximation and projection (UMAP) with hierarchy density-based spatial clustering of applications with noise (HDBSCAN). This approach effectively generates low-dimensional configurational embeddings. The hierarchical application of this architecture, in conjunction with MSM, reveals global kinetic connectivity while identifying local conformational states. Consequently, our methodology establishes a multiscale mechanistic elucidation framework. Leveraging the benefits of the uniform sample distribution and a denoising approach, our model demonstrates robustness in preserving global and local data structures compared to traditional dimension reduction methods in the field of MD analysis area. The interpretability of hyperparameter selection and compatibility with downstream tasks are cross-validated across various simulation data sets, utilizing both computational evaluation metrics and experimental kinetic observables. Furthermore, the predicted Mcl1-BH3 association kinetics (0.76 s-1) is in close agreement with surface plasmon resonance experiments (0.12 s-1), affirming the plausibility of the identified pathway composed of representative conformations. We anticipate that the devised workflow will serve as a foundational framework for studying recognition patterns in complex biological systems. Its contributions extend to the exploration of protein functional dynamics and rational drug design, offering a potent avenue for advancing research in these domains.

Tracer-test-based dimensionality reduction model for characterizing fracture network and predicting flow and transport in fracture aquifer

Dimensionality reduction models (DRM) are widely utilized for well-posed inversion of highly heterogeneous aquifer properties. However, these models predominantly rely on predefined aquifer properties, which are often unknown in realistic aquifers; moreover, DRM for fracture network characterization is still lacking. In this study, we develop a novel DRM based on the tracer breakthrough curve (TBC) obtained from cross-hole tracer test. This model enables integrated characterization of fracture networks and prediction of flow and transport in fractured aquifers. Firstly, the relationship between TBC shape and flow and transport in fracture networks is elucidated through 3D-printing physical modeling and numerical simulations. It is observed that in complex fracture networks, flow and transport dominantly occur through a limited number of preferential flow paths under pumping scenarios involving simultaneous injection and extraction. The preferential flow paths remain stable under different pumping rates, with their number aligning with the number of inflection points, including concentration peaks in TBC and evident slowness point during the decline of tracer concentration. Building upon this finding, DRM composed of parallel fractures with the number of fractures equal to the number of inflection points in TBC is constructed. Parallel fractures in DRM represent preferential flow paths for flow and transport within the original fracture network. Parameters including sizes, apertures, and separation of parallel fractures are then inferred by TBC inversion. It is tested by the 3D-printing physical model that under pumping rates differential to that implemented in the tracer test for DRM parameters estimation, the established DRM is capable of predicting the tracer concentration and outflow temperature at the extraction well with a relative error lower than 10%. Moreover, the implementation of proposed DRM concept with synthetic large fracture network demonstrates its genericity for the characterization of fracture networks at field scale. This new DRM concept provides a feasible and effective way for fracture network characterization and flow and transport prediction.

Disentangling the response of vegetation dynamics to natural and anthropogenic drivers over the Qinghai-Tibet Plateau using dimensionality reduction and structural equation model

The Qinghai-Tibet Plateau (QTP), also referred to as the “roof of the world”, is a distinct alpine ecosystem located at an average elevation of more than 4000 m. Climate variation and anthropogenic activity have affected vegetation development on the QTP during the past few decades, but it is still debatable how significant a role any of these causes has played. According to land surface air temperature data around the world, during the past 50 years, temperatures on the QTP have risen twice as rapidly as the average global level. Previous studies have employed the residual method to examine the impacts of anthropogenic activities on vegetation development, with a lack of actual data validation and less attention to explaining the complex interactions and pathways between vegetation growth, natural factors (climate change and topography), and anthropogenic drivers. Furthermore, although the relationships between vegetation dynamics and terrain have been addressed, the interactions in which meteorological factors influence vegetation growth in accordance with topographical variations remain inadequately elucidated. The raster datasets employed in this study include MODIS NDVI, meteorology, topography, as well as anthropogenic activity from 2000 to 2018. A suite of methods including partial correlation coefficient (PCC), principal component analysis (PCA), and partial least squares structural equation model (PLS-SEM) were applied to explore the interrelationships between various influencing factors and regional vegetation development over the past two decades. The results suggested that: (1) an elevation-specific temperature threshold existed, indicating that temperature could promote or inhibit vegetation growth depending on whether it is below or above this threshold, respectively, which was controlled by a trendline equation; (2) the PLS-SEM and PCA results showed that vegetation development was dominantly promoted by temperature and was most sensitive to precipitation; (3) topography, especially elevation, affected vegetation growth indirectly by influencing climatic change, while anthropogenic drivers had positively contributed to vegetation dynamics on the QTP under the influence of implementing ecological management such as the nature reserves establishment.

Statistical Analysis of Complex Shape Graphs.

This paper provides developments in statistical shape analysis of shape graphs, and demonstrates them using such complex objects as Retinal Blood Vessel (RBV) networks and neurons. The shape graphs are represented by sets of nodes and edges (articulated curves) connecting some nodes. The goals are to utilize nodes (locations, connectivity) and edges (edge weights and shapes) to: (1) characterize shapes, (2) quantify shape differences, and (3) model statistical variability. We develop a mathematical representation, elastic Riemannian metrics, and associated tools for shape graphs. Specifically, we derive tools for shape graph registration, geodesics, statistical summaries, shape modeling, and shape synthesis. Geodesics are convenient for visualizing optimal deformations, and PCA helps in dimension reduction and statistical modeling. One key challenge in comparing shape graphs with vastly different complexities (in number of nodes and edges). This paper introduces a novel multi-scale representation to handle this challenge. Using the notions of (1) "effective resistance" to cluster nodes and (2) elastic shape averaging of edge curves, it reduces graph complexity while retaining overall structures. This allows shape comparisons by bringing graphs to similar complexities. We demonstrate these ideas on 2D RBV networks from the STARE and DRIVE databases and 3D neurons from the NeuroMorpho database.

Unsupervised Dimensionality Reduction Modeling for Analyzing Aging Characteristics of Hanji

Unsupervised Dimensionality Reduction Modeling for Analyzing Aging Characteristics of Hanji

Comparison of dimensionality reduction methods and evaluation of prediction effect based on cardiovascular disease data

Cardiovascular diseases have become the leading cause of death in the world, and the mortality rate caused by them is still on the rise, which is a major challenge facing the world. In the study of cardiovascular diseases, there are many kinds and quantities of factors leading to the disease, so how to screen more effective factors for research and accurate prediction is an important problem. The aim of this study is to conduct an in-depth comparative analysis of multiple dimensionality reduction methods for cardiovascular disease data. This paper evaluates the results produced by various dimensionality reduction methods and models, and uses Accuracy Rate, Confusion Matrix and Area Under Curve (AUC) and other indicators to evaluate their prediction effects. The study finds that the decision tree model has the best performance and is superior to LDA linear discriminant analysis in terms of accuracy and data dimension. The principal component analysis method is relatively complex, although it can effectively reduce the data dimension, its accuracy in forecasting decreases, and this method is not conducive to horizontal and vertical comparison and the accumulation of statistical data.

The interaction of obesity with susceptible gene polymorphisms in the relationship with mild cognitive impairment.

Mild cognitive impairment (MCI) in the elderly is threatening the mental health of the elderly, and the interaction of some factors is worth exploring. This study aims to explore the interactions of obesity and gene polymorphisms in the relationship with MCI. A total of 2555 community resident dwellings include 444 participants who met MCI criteria recruited from the Ningxia province of China. Fourteen MCI-susceptible single nucleotide polymorphisms were detected using a high-throughput mass spectrometer. The interaction was examined by performing the multifactor dimensionality reduction model and unconditional logistic regression model. Logistic regression showed that obesity (OR = 1.42, 95%CI: 1.04-1.94), rs2075650G allele carrying (OR = 17.95, 95%CI: 1.32-244.95), rs11556505T allele carrying (OR = 0.06, 95%CI: 0.01-0.87) were statistically associated with MCI. Multifactor dimensionality reduction analysis showed a strong antagonistic effect between obesity and rs4402960 (Interaction dendrogram between obesity and rs4402960 is red) and a weak synergy effect on rs7901695 (Interaction dendrogram between obesity and rs7901695 is green). The hierarchical analysis showed obesity is a risk factor for MCI in the non-rs4402960T allele carrier group (OR = 1.55, 95%CI: 1.02-2.35). This study found that obesity is an independent risk factor for MCI, and the interactions with MCI-susceptible gene polymorphisms suggest a possible precision preventive intervention program should be developed to reduce the risk of MCI among individuals with obesity in the community.

Sex-Specific Perturbation of Systemic Lipidomic Profile in Newborn Lambs Impacted by Prenatal Testosterone Excess.

Gestational hyperandrogenism adversely impacts offspring health. Using an ovine model, we found that prenatal testosterone (T) excess adversely affects growth and cardiometabolic outcomes in female offspring and produces sex-specific effects on fetal myocardium. Since lipids are essential to cardiometabolic function, we hypothesized that prenatal T excess leads to sex-specific disruptions in lipid metabolism at birth. Shotgun lipidomics was performed on the plasma samples collected 48 hours after birth from female (F) and male (M) lambs of control (C) and (T) sheep (CF = 4, TF = 7, CM = 5, TM = 10) and data were analyzed by univariate analysis, multivariate dimensionality reduction modeling followed by functional enrichment, and pathway analyses. Biosynthesis of phosphatidylserine was the major pathway responsible for sex differences in controls. Unsupervised and supervised models showed separation between C and T in both sexes with glycerophospholipids and glycerolipids classes being responsible for the sex differences between C and T. T excess increased cholesterol in females while decreasing phosphatidylcholine levels in male lambs. Specifically, T excess: 1) suppressed the phosphatidylethanolamine N-methyltransferase (PEMT) phosphatidylcholine synthesis pathway overall and in TM lambs as opposed to suppression of carnitine levels overall and TF lambs; and 2) activated biosynthesis of ether-linked (O-)phosphatidylethanolamine and O-phosphatidylcholine from O-diacylglycerol overall and in TF lambs. Higher cholesterol levels could underlie adverse cardiometabolic outcomes in TF lambs, whereas suppressed PEMT pathway in TM lambs could lead to endoplasmic reticulum stress and defective lipid transport. These novel findings point to sex-specific effects of prenatal T excess on lipid metabolism in newborn lambs, a precocial ovine model of translational relevance.

Non-destructive prediction of TVB-N using color-texture features of UV-induced fluorescence image for freeze-thaw treated frozen-whole-round tilapia.

The investigation of UV-induced fluorescence imaging coupled with machine learning was conducted to non-destructively detect the total volatile basic nitrogen (TVB-N) of frozen-whole-round tilapia (FWRT) during freezing and thawing. The UV-induced fluorescence images of FWRT at the wavelength of 365 nm were acquired by self-developed fluorescence image acquisition system. In total, 169 color and texture features based on RGB, hue-saturation-intensity and L*a*b* color spaces and gray level co-occurrence matrix were extracted, respectively. Successive projections algorithm (SPA) was employed to select the optimal 16 features to achieve feature dimension reduction modeling. With full and extracted features as input, the models of partial least squares regression (PLSR), least-squares support vector machine (LSSVM) and convolutional neural network (CNN) were established for TVB-N prediction. Results indicated that the full features-based CNN performed better than SPA based prediction models (SPA-PLSR and SPA-LSSVM). The CNN model was determined to be the optimal with an RP2 value of 0.9779, RMSEP value of 1.1502 × 10-2 g N kg-1 and RPD value of 6.721 for TVB-N content predictiin. The CNN method based on UV fluorescence imaging technology has potential for quality and safety detection of FWRT. © 2023 Society of Chemical Industry.