Sleep architecture, including spindles and slow oscillations, is disrupted in Alzheimer disease (AD). How changes in these sleep elements relate to cognitive decline is less clear. Our objectives were to examine changes in sleep macroarchitecture and microarchitecture in early clinical stages of AD compared with older adult controls (CTLs) and to investigate their associations with longitudinal cognitive change. This was both a cross-sectional and longitudinal study performed at Mass General Brigham Hospitals, where participants with early AD or CTLs underwent overnight ambulatory scalp EEG and longitudinal cognitive testing. We compared sleep microarchitectural features extracted from the EEG, including spindle activity, across the brain topography and between groups. We then performed longitudinal analyses using mixed-effects models to test the association of these sleep features with changes in cognition on the Montreal Cognitive Assessment (MoCA), collected annually for up to 7 years. AD (n = 47, mean age 74.1 years, 66% female) and CTL (n = 43, mean age 72.6 years, 56% female) groups spent a similar proportion of sleep time in each stage of sleep. Sleep efficiency, however, was lower in the AD group (mean: CTL 75.1% vs AD 70.9%; p = 0.034). We found a significant reduction in spindle range power (11-16 Hz) in patients with AD compared with CTLs, particularly in the temporal regions (mean normalized power at EEG channels T3/T4: CTL 3.13 ± 1.13 vs AD 2.48 ± 1.01; p = 0.005). In participants with longitudinal MoCA scores (AD = 26, CTL = 25), reduced temporal lobe spindle density (β = 0.61, 95% CI 0.35-0.87; false discovery rate [FDR]-adjusted p < 0.001) and temporal lobe spindle power (β = 0.56, 95% CI 0.22-0.88; FDR-adjusted p = 0.005) were each associated with a faster rate of cognitive decline. Temporal lobe sleep spindle activity is reduced in early clinical stages of AD and is associated with a faster rate of cognitive decline. Our results underscore the importance of including temporal lobe measurements when assessing sleep neurophysiology in AD, which is not standard in polysomnography. Future work examining the relationship between AD biomarkers and reduced spindle activity is needed to elucidate the potential mechanisms underlying these findings.

Comparative performance of Scribe and database search engines in metaproteomic profiling of a ground-truth microbiome dataset.

Long-term exposure to traffic-related air pollution is associated with epigenetic age acceleration.

Cortical thickness alterations in afro-descendants with schizophrenia and bipolar disorder: An exploratory analysis.

Objective Chronic obstructive pulmonary disease (COPD) remains a leading cause of disability and mortality among elderly populations. Studies indicate that AGER plays a critical regulatory role in the pathogenesis of respiratory disorders. However, the genetic variations in AGER to COPD susceptibility remain incompletely understood. This study employs a case–control design to investigate associations between AGER genetic variants and COPD risk in the Southern Chinese Han population. Methods This study enrolled 270 COPD patients and 271 healthy controls. AGER single-nucleotide polymorphisms (SNPs) were analysed using the MassARRAY iPLEX platform. Logistic regression models evaluated associations between AGER polymorphisms and COPD susceptibility, with false discovery rate (FDR) correction applied to mitigate multiple testing errors. SNP–SNP interactions were investigated through multifactor dimensionality reduction (MDR) analysis. Expression quantitative trait locus (eQTL) data from the GTEx database were further analysed to assess regulatory relationships between SNPs and AGER gene expression levels. Results This study showed that rs3134941 (G allele, OR = 0.21, 95% CI = 0.10–0.41, p (FDR) = 0.001) and rs3131300 (G allele, OR = 0.32, 95% CI = 0.20–0.49, p (FDR) = 0.0001) were significantly associated with a reduced susceptibility to COPD. MDR indicated that rs3131300 was the optimal predictive model for COPD risk. Additionally, initial mechanistic investigations utilizing the GTEx database identify rs3134941 (C > G) and rs3131300 (A > G) as significant expression quantitative trait loci for AGER mRNA in cell-cultured fibroblasts and whole blood. Conclusion Our study demonstrated that AGER genetic variants might play a protective role in the progression of COPD.

Early detection of knee osteoarthritis (OA) is important. Spin-lattice relaxation in the rotating frame (T1ρ) mapping is sensitive to early cartilage changes, but the mono-exponential (ME) model may be limited. Multi-component models can capture more tissue complexity, but their diagnostic advantage has not been validated. To evaluate if stretched- (SE) and bi-exponential (BE) T1ρ models can improve early knee OA detection over the ME model. Case-control study. Twenty-six healthy subjects (mean age 51.5) and 26 early knee OA patients (mean age 61.8). T1ρ-prepared Turbo FLASH sequence at 3 T field strength. T1ρ parameters from three exponential models were adjusted for age. To maximize group separability, the parameters were combined into single discriminators for both global knee cartilage and six anatomical sub-regions. Diagnostic performance was assessed based on the ability of these combined models to distinguish early OA. Parameters were adjusted for age. Mann-Whitney U-test (group comparisons), linear discriminant analysis (LDA), and area under the receiver operating characteristic (ROC) curve (AUC) with bootstrapped 95% confidence intervals (CI). Significance level set at p < 0.05, using the false discovery rate (FDR) to correct for multiple comparisons. In the global analysis, no model demonstrated significant diagnostic performance (p-values of 0.63, 0.96, 0.63 for ME, SE, and BE). Multi-regional SE model (AUC = 0.83, CI: 0.72, 0.93) significantly distinguished OA and healthy groups. Calibration analysis showed the SE model had the lowest Brier score (0.17), significantly better than the ME model (0.26). Sub-regional analysis of T1ρ parameter maps suggests an improvement in diagnostic performance for early knee OA compared to globally averaged measurements. The stretched-exponential model showed the most promise. However, small sample size and wide confidence intervals highlight the need for further validation with a larger cohort before clinical utility claims can be made. 4. Stage 2.

Postoperative delirium (POD) represents a significant challenge in perioperative care, particularly among older surgical patients. This acute neuropsychiatric syndrome is associated with prolonged hospitalization, increased mortality, and long-term cognitive decline. Sleep disturbance has emerged as a significant modifiable risk factor for POD. Sodium oxybate (SO), a gamma-aminobutyric acid B (GABAB) receptor agonist with established sleep-enhancing properties, presents a promising therapeutic approach for POD prevention. The objective of this trial was to investigate whether prophylactic intraoperative sodium oxybate reduces POD incidence in older patients (≥ 65years) undergoing major orthopedic surgery. This randomized, double-blind, placebo-controlled trial enrolled 332 older patients undergoing elective spine and joint replacement surgery. Participants received either sodium oxybate (30mgkg-1) or saline after anesthetic induction. Stratified randomization allocated equal numbers to morning and afternoon surgery groups. The primary outcome was POD incidence within seven postoperative days, assessed using the Confusion Assessment Method (CAM). POD incidence showed no significant difference between groups in unstratified population (10.3% vs. 13.5%, P = 0.372). However, subgroup analysis revealed protective effects in morning surgery patients (7.3% vs. 18.5%, relative risk (RR) = 0.395, 95% confidence intervals (CI) = 0.161-0.968, P = 0.033), while no effect was observed in the afternoon surgery group (13.3% vs. 8.5%, P = 0.318). Among patients with delirium, no significant differences were observed in delirium severity, onset timing, delirium duration, or subtype distribution after false discovery rate (FDR) correction. No significant differences were found in sleep quality, maximal pain score, or safety parameters between groups after FDR correction. Intraoperative sodium oxybate demonstrates possible time-specific efficacy, significantly reducing POD incidence exclusively in older patients undergoing morning orthopedic surgery, while demonstrating an acceptable safety profile with no significant adverse effects on anesthesia recovery or hemodynamic parameters, suggesting a potential chronotherapeutic approach to POD prevention. Chinese Clinical Trial Registry, ChiCTR2300078594. Registered on 2023-12-13.

This study aimed to investigate the causal relationship between dietary habits and nine common diseases with high female prevalence using a two-sample Mendelian randomization (MR) approach. Drawing upon pooled Genome-Wide Association Study (GWAS) data from European cohorts, this research interrogated genetic associations across 21 detailed dietary classifications and nine major diseases with high prevalence in women. Our study identified multiple associations between dietary factors and gynecological disease risks. In initial analyses, dried fruit was inversely associated with breast and cervical cancer, while beef increased breast cancer risk. Endometrial cancer risk was linked to coffee, alcohol, and non-oily fish. Ovarian cancer risk exhibited mixed associations, rising with alcohol and non-oily fish but decreasing with pork, dried fruit, and beef. Among benign conditions, risk associations were observed for ovarian cysts (positive with omelette/whole-wheat; negative with coffee/lobster/dark chocolate/dried fruit), endometriosis (positive with shellfish; negative with coffee/cheese), uterine fibroids (coffee, dried fruit, non-oily fish), premature ovarian failure (protective: soya dessert; risk: herbal tea, dark chocolate, whole-wheat), and polycystic ovary syndrome (fresh fruit, yogurt, herbal tea). However, after false discovery rate (FDR) correction for multiple testing, only two associations remained significant: dried fruit intake with a reduced risk of breast cancer and coffee intake with an increased risk of uterine fibroids. Our comprehensive MR analysis yields insights into the potential causal links between dietary factors and common diseases in women, paving the way for non-pharmacological public health interventions. The findings highlight the potential of dietary modifications as a preventative measure against the onset of these conditions.

Cross-linking mass spectrometry has become a powerful technique for identifying protein interactomes and studying protein structures. We report the development of homo-bi-functional photo-activatable (BFPA) cross-linkers for cross-linking mass spectrometry. These cross-linkers theoretically react with any-to-any amino acid, overcoming limitations in amino acid reactivity. Different fragmentation energies and cross-linking conditions were tested, and the false discovery rate was benchmarked against a non-cross-linked sample and a false protein sequence search across different search engines. BFPA cross-linkers identified cross-links with better overall agreement with high resolution protein structures compared to randomized cross-links and compared to other commonly used cross-linkers. Different cross-link sites were identified by BFPA cross-linkers across bovine serum albumin, human importin complex and dengue protein complex, demonstrating its wide applicability to different protein complexes. BFPA have the potential to serve as complementary tools to current cross-linkers to expand the coverage of cross-linking mass spectrometry experiments.

PurposeInflammation is implicated in the pathogenesis of gastroenteropancreatic neuroendocrine neoplasms (GEP‐NENs); however, the causal nature of this association remains unclear. This study sought to evaluate the causal relationships between GEP‐NENs and inflammatory factors using a two‐sample Mendelian randomization (MR) approach.MethodsWe performed a two‐sample MR analysis to investigate the causal associations between 91 inflammatory proteins and 731 immune cell traits as exposures and the five subtypes of GEP‐NENs as the outcomes. The analytical approach employed various methodologies, such as inverse variance weighting, MR‐Egger, weighted mode, weighted median, and simple mode. To evaluate the robustness of the results, sensitivity analyses were conducted, which encompassed MR Egger regression, MR multiple gene residual and outlier detection, leave‐one‐out analysis, and Cochran’s Q test. False discovery rate (FDR) correction was applied, and causal relationships at the gene level were deemed significant at p < 0.05 after FDR adjustment.ResultsAfter FDR correction, the findings revealed robust causal associations between genetically predicted HLA DR++ monocyte %leukocyte (OR = 3.09, 95% CI: 1.76–5.44, p < 0.001, FDR = 0.022), HLA DR on CD14+ CD16‐ monocyte (OR = 1.72, 95% CI: 1.34–2.22, p < 0.001, FDR = 0.010), and HLA DR on CD14+ monocyte (OR = 1.76, 95% CI: 1.36–2.29, p < 0.001, FDR = 0.010) and genetically predicted stomach NENs. Reverse analysis revealed that GEP‐NENs had no major impact on inflammation.ConclusionThese findings reveal the immune mechanisms underlying GEP‐NENs and highlight potential therapeutic strategies targeting the immune microenvironment of GEP‐NENs.

BackgroundSubjective cognitive complaints (SCCs) can be an early indicator of Alzheimer's disease and related dementias. SCCs have been shown to be common in people exposed repetitive head impacts (RHI), particularly male former professional American football players. This study characterized participant and informant‐reported SCCs in terms of rate, concordance with standardized neuropsychological measures, and potential associated factors among participants with diverse sources/severity of RHI exposure.MethodThe sample included participants with (n = 172) and without (n = 320) RHI from the Boston University Alzheimer's Disease Research Center Clinical Core. RHI status is based on the 2021 NINDS TES Research Diagnostic Criteria. The Cognitive Change Index (CCI) and BRIEF‐A Meta‐Cognition Index (MI) measured self and informant‐reported SCCs. Participants completed neuropsychological assessments of memory (Craft Story 21 Recall, NAB List Learning Long Delay) and executive function (Trails B). Informants completed the Neuropsychiatric Inventory‐Questionnaire (NPI‐Q). ANCOVAs compared performance of RHI/non‐RHI groups on the SCC measures. Pearson correlation examined agreement between participant/informant responses. Multivariable linear regression models tested associations between SCCs and neuropsychological tests and examined correlates of SCCs. All models controlled for age, sex, race, and education. p‐values were false discovery rate adjusted.ResultTable 1 describes the sample. Compared to non‐RHI, the RHI group was younger, likelier to be male, and likelier to have MCI. The RHI group had significantly higher MI (B=8.084, p = 0.006), informant MI (B=9.014, p = 0.006), CCI (B=5.986, p <0.001), and informant CCI scores (B=7.062, p <0.001. Self/informant CCI and MI scores were more correlated in the RHI (r = 0.592, 0.540, respectively) vs non‐RHI group (r = 0.416, 0.373, respectively). Figure 1. Within the RHI group, there were associations between participant/informant SCCs and the objective measures (e.g., B=‐0.086, padj<0.001 for CCI and NAB) (Table 2). We observed fewer, weaker associations between SCCs and neuropsychological measures in the non‐RHI group. NPI‐Q was a consistent correlate of self/informant SCCs. Demographics (e.g., self/informant race) were also associated but to a lesser extent.ConclusionIn RHI settings, SCCs might be more frequent and reflect cognitive function. High SCC rates are likely multifactorial, with influence from neuropsychiatric factors. Future research should examine longitudinal change in self/informant SCCs and correlation with disease biomarkers.

Gut dysbiosis and aberrant immune activation are increasingly recognized as critical determinants of acute respiratory distress syndrome (ARDS). However, the causal contributions of specific gut taxa, immune-cell phenotypes, and their interactive pathways remain incompletely understood. In this study, we conducted a comprehensive two-sample Mendelian randomization (MR) analysis to elucidate the individual and combined effects of the gut microbiome and immune milieu on ARDS susceptibility. Using 5 combined methodologies, the primary causal estimates were primarily derived through the Inverse-Variance Weighted approach. Heterogeneity was evaluated using Cochrane’s Q test, while horizontal pleiotropy was assessed via the MR-Egger intercept, and robustness was confirmed through leave-one-out and reverse MR analyses. Following adjustments for the false discovery rate (FDR), our findings indicated that, although the overall effects of exposures on ARDS were not statistically significant (PFDR < 0.2), there were causal associations identified for 12 gut microbiota taxa, 24 immune cells, and 6 circulating inflammatory cytokines with ARDS (P < .05). Initial mediation analyses indicated that EIF4EBP1, caspase-8, IL-6, and IL-8 might partly mediate these effects, but 1000 BCa bootstrap iterations rejected all indirect pathways. These findings underscore the pivotal roles of gut microbiota and immune factors, both individually and interactively, in the pathogenesis of ARDS, offering a genetically informed basis for future treatments targeting the microbiome and immune system.

BackgroundHealth determinants such as income and employment significantly influence declarative memory in women with HIV (WWH). These social exposures may contribute to cognitive decline through complex biological mechanisms. We used high‐throughput metabolomics data and causal mediation analysis to identify multiple biological pathways linking adverse socioeconomic conditions and cognitive decline.MethodA total of 324 women (n = 225 WWH) from the New York sites of the Women's Interagency HIV Study completed biennial cognitive testing, starting in 2009 and every two years, which included the Hopkins Verbal Learning Test (HVLT)‐R, a measure of verbal learning and memory. Metabolomic assessment was conducted from serum samples collected prior to or at first cognitive visit using liquid chromatography tandem mass spectrometry. We used multivariable generalized estimating equation models to link peripheral metabolomic profiles and cognitive trajectories in verbal learning and memory. Linear mixed effect models were used to predict subject‐specific trajectories for downstream mediation models. We used regression‐based multiple mediation analyses to estimate the joint effect of multiple metabolic mediators as pathways between income, employment, and cognitive trajectories. We considered a false discovery rate (FDR) of 5% for metabolomic discovery.ResultOur sample had average age of 43.1 years and were followed‐up for 11.9 years, 63% were non‐Hispanic Black, 76% reported earnings of <24K, and 65% were unemployed. We identified 16 FDR‐corrected metabolites associated with trajectories in verbal memory and 15 with verbal learning. Top metabolites (serotonin, taurine, adenosine, niacinamide, α‐glycerophosphocholine, ADMA, pseudouridine, and sphingomyelins) were common in both domains. Stratified models by serostatus indicated similar results in WWH and women without HIV. In causal mediation analysis, we found that 30% (95%CI: 4%‐100%) of the effect of employment and 18.2% (95%CI: ‐2.1% to 72.3%) of the effect of income on verbal learning was mediated through differences in metabolite levels. We also found that 35.6% (2.7%‐100%) of the effect of employment and 25.4% (95%CI: ‐1.7%‐100%) of the effect of income on verbal memory was mediated through altered metabolomic signatures.ConclusionHealth determinants influence biological process implicated in cognitive decline. Biosocial interventions are needed to address the conditions that perpetuate poor cognitive health in women.

We characterised visual field (VF) spatial loss patterns in acute non-arteritic anterior ischaemic optic neuropathy (NAION) using archetypal analysis (AA). We performed standard automated perimetry on 727 participants with acute NAION at screening, day 1 of enrolment, months 2, 6 and 12. We applied AA, an unsupervised machine learning technique, to identify and quantify distinct VF loss patterns (archetypes, ATs). We used Mann-Whitney U tests and Wilcoxon signed-rank tests to assess demographic differences and longitudinal changes in discrete variables, and linear regression to analyse continuous variables. AA identified 10 distinct ATs, with three especially prevalent patterns: global VF loss (AT1), inferior altitudinal loss (AT2) and inferior altitudinal with macular sparing (AT3). Between day 1 and month 2, AT1 RW significantly increased from 15.7% to 28.1% (false discovery rate, FDR-adjusted p<0.001). Asian participants consistently exhibited greater RW for AT1 compared with White participants (day 1: 40.7% vs 17.7%; FDR-adjusted p<0.001). Sex differences emerged modestly at months 2 and 6, with females having a higher RW for superior altitudinal loss (FDR-adjusted p=0.049). Older participants showed slightly greater frequency of central horizontal and mild central depression patterns (FDR-adjusted p<0.001). AA effectively quantifies distinct, clinically significant VF spatial loss patterns in NAION, revealing significant temporal changes and demographic differences. Global VF loss represents the predominant AT, increasing notably within 2 months of disease onset. Prominent racial disparities, particularly higher severity in Asian individuals, underscore potential differences in NAION aetiology or susceptibility. These findings provide a foundation for improved disease characterisation and prognosis.

ABSTRACTIntroductionAlcohol marketing significantly influences consumption patterns, particularly among youth, heavy drinkers and women. Digital platforms have amplified this impact through targeted, immersive campaigns. However, monitoring such marketing remains a challenge due to its opaque and dynamic nature. This study introduces SCANNER Alcohol, an AI‐enabled system designed to detect alcohol marketing in online content at the brand level.MethodsSCANNER Alcohol integrates object detection (logos) and optical character recognition (text) to automatically identify 134 alcohol brands within image and video data. The system was trained using annotated datasets of brand logos and validated through standard machine learning metrics. Real‐world performance was assessed using social media screen recordings (brand accounts and 12 h of general use), benchmarked against manual coding.ResultsSCANNER Alcohol achieved high algorithmic accuracy, with a mean average precision of 0.94, recall of 0.96 and F1 score of 0.95. In real‐world testing, the model demonstrated strong performance, correctly identifying 98.9% of alcohol‐branded posts in social media videos. SCANNER Alcohol achieved a low 6.7% false discovery rate, indicating high precision and low noise in the detection output.Discussion and ConclusionsSCANNER Alcohol is the first system to combine brand‐level object and text detection for automated digital alcohol marketing surveillance. Its high accuracy on real‐world data and ethical design make it a valuable tool for public health monitoring. SCANNER Alcohol offers a scalable, adaptable and ethically sensitive solution to support regulatory efforts to hold alcohol companies accountable for their digital marketing practices and to ultimately reduce alcohol‐related harm.

BackgroundLimbic‐predominant age‐related TDP‐43 encephalopathy neuropathological change (LATE‐NC) is the third most significant contributor to late‐onset dementia after Alzheimer's disease neuropathologic change (ADNC; amyloid‐β and tau) and cerebrovascular disease. However, the molecular alterations underlying LATE‐NC remain poorly understood.MethodWe studied 959 participants (Age=89.4±6.7, Female: 65.9%) from Religious Orders Study and the Rush Memory and Aging Project (ROSMAP). Bulk RNA sequencing (RNA‐seq) was performed on post‐mortem brain tissues from the amygdala (n = 97) and dorsolateral prefrontal cortex (DLPFC; n = 937), with 75 overlapping participants. Linear regression models were used to evaluate associations between RNA expression and LATE‐NC burden (TDP‐43 immunohistochemistry; the average semi‐quantitative counts of neuronal cytoplasmic inclusions across six brain regions), adjusting for age, sex, and quantitative ADNC burden (amyloid‐β and tau across eight brain regions). Gene Ontology (GO) terminology overrepresentation was assessed for genes with positive or negative associations with LATE‐NC.ResultsRNA‐seq analysis of the amygdala (n = 97) identified 257 genes positively correlated and 178 genes negatively correlated with LATE‐NC at false discovery rate (FDR) <0.05 (Figure 1A). Downregulated pathways (Figure 1B) included regulation of membrane potential, synaptic transmission, and vesicle‐mediated transport, while upregulated pathways (Figure 1C) involved microtubule‐based movement and cilium assembly. By contrast, despite a much larger sample size, we did not observe any associated genes in the DLPFC (all FDR>0.05)—a region involved only in advanced LATE‐NC. Targeted analysis of the amygdala LATE‐NC‐associated genes in the DLPFC (nominal p <0.05, with a consistent direction of association) revealed 8 positively associated and 28 negatively associated genes (Figure 2A‐C). The LATE‐NC‐associated genes downregulated in both amygdala and DLPFC captured synaptic function and vesicle transport pathways, reflecting shared neuronal/synaptic disruption across brain regions in LATE‐NC.ConclusionsOur findings highlight distinct molecular signatures of LATE‐NC in the amygdala, including upregulation of microtubule‐related pathways and downregulation of neuronal/synaptic genes. Interestingly, this LATE‐NC signature was much weaker in the neocortex, underscoring the critical importance of brain‐region‐specific multi‐omic studies in neurodegeneration. Our results provide one of the first large‐scale molecular atlas of LATE‐NC from the limbic region, providing foundations for disease models, biomarker discovery, and therapeutic approaches.

Electroencephalogram (EEG) plays a vital role in seizure detection, yet existing methods often fail to adequately capture the spatiotemporal characteristics of EEG signals, leading to limited performance. Moreover, most current models depend on supervised learning and thus require large amounts of labeled data. To address these issues, this paper introduces the Long Short-Term Memory-Transformer (LTformer) encoder, designed to model long-term temporal dependencies in EEG signals while retaining spatial information across electrode channels. We further propose a dual-stream self-supervised learning (SSL) strategy to pretrain the model, enabling the LTformer encoder to learn discriminative representations from extensive unlabeled EEG data. After pretext training, the encoder is transferred and fine-tuned for downstream seizure detection. The proposed method, termed Self-Supervised Attention LTformer (SALT), is evaluated on two public EEG datasets using both segment-based and event-based experimental protocols. In the segment-based evaluation, SALT achieves 98.87% sensitivity, 99.15% accuracy, and 99.41% specificity on CHB-MIT, and 98.04% sensitivity, 97.72% accuracy, and 97.62% specificity on Siena. In the event-based evaluation, SALT attains 98.57% sensitivity with a false discovery rate (FDR) of 0.26 on CHB-MIT, and 98.65% sensitivity with an FDR of 0.25 on Siena. The code is publicly available at https://github.com/peutim114/SALT .

BackgroundTranscriptome‐wide association studies (TWAS) extend Genome‐wide association studies by integrating gene expression data to explore the functional consequences of non‐coding variants. This study employs a multi‐ancestry TWAS approach to identify AD‐associated genes across groups, leveraging data from the Multi‐Ancestry Genomics, Epigenomics, and Transcriptomics of Alzheimer's (MAGENTA) Project. The study includes 224 African American ancestry (AFA), 235 European ancestry (EUR), and 298 Hispanic ancestry (HIS) individuals with both whole‐blood expression and genotype data.MethodWe first analyzed the MAGENTA data to identify specific genetic variants that impact gene expression across the three groups using the Sum of Shared Single Effects (SuShie) method. We examined a region of 500 kilobases around each gene and SNPs with a minor allele frequency > 0.01. SuShiE uses a sparse modeling assumption, assuming no effects at most SNPs. This analysis produces credible sets which are highly likely to contain variants causally associated to gene expression. The study adjusted for covariates including age at exam, sex, AD status, principal components, cell‐type proportions, and PEER factors in the cis‐eQTL weight prediction model. We also applied MA‐FOCUS (Multi‐Ancestry Fine‐Mapping of Causal Gene Sets) to fine‐map TWAS associations at genomic risk regions with at least one significant gene.ResultOur analysis identified five AD‐related genes in three regions. Among those identified, the AD‐related TREML2 was observed across all three groups. The SSW meta‐analysis of ancestry‐specific TWAS associations identified TREML2 with a false discovery rate (FDR)<0.1. The gene's effect size directions were positive and consistent across all ancestries (SSW Z = 4.14; EUR Z=3.76; AFR Z=2.34; HISP Z=0.97), with a MA‐FOCUS PIP (posterior inclusion probability) of 0.88, suggesting robust association.ConclusionThis study highlights the importance of multi‐ancestry approaches in uncovering the genetic mechanisms influencing the molecular pathways involved in AD development. Zheng et al. found that the elevated TREML2 expression might exert pro‐inflammatory and proliferation promoting effects on microglia in an AD context. Our findings underscore the potential of TREML2 as a target for further research and therapeutic intervention in AD. Increased sample sizes in studied populations will advance the identification of more AD‐related genes.

Abstract In high-stakes scientific contexts, explainable AI is crucial for deriving meaningful insights from complex tabular data. A formidable challenge is ensuring both rigorous statistical guarantees and clear interpretability in feature extraction. While traditional methods like PCA are limited by linear assumptions , powerful neural network approaches often lack the transparency required in scientific domains. To address this gap, we introduce \TheName{}, a novel self-supervised learning pipeline that makes nonlinear feature interactions interpretable. \TheName{} marries the power of kernel principal components for capturing complex dependencies with a sparse, principled polynomial representation to achieve clear interpretability with statistical rigor. Our approach bridges data-driven complexity and statistical reliability via three stages. First, it generates self-supervised signals using kernel principal components to model complex patterns. Second, it distills these signals into sparse polynomial functions for interpretability. Third, it applies a multi-objective knockoff selection procedure with significance testing to rigorously identify important features while controlling the false discovery rate (FDR). Extensive experiments on diverse real-world datasets demonstrate the effectiveness of \TheName{}, consistently surpassing other methods in feature selection for regression and classification tasks. In particular, applications on physics datasets highlight the ability of the proposed method to produce scientifically valid and interpretable explanations, reinforcing its practical utility and the critical role of explainability in AI for science.

BackgroundTraditional training methods often fall short in replicating the perceptual load of match environments. Virtual reality (VR) has emerged as a promising modality to enhance cognitive-motor integration in football contexts. Purpose: This study aimed to examine the effects of a 7-week virtual reality-based training program designed to improve both scanning behavior and passing performance in youth football players, in comparison with traditional training methods. Methods: A parallel-group randomized controlled trial was conducted with 22 male youth players from U16–U17 squads (mean age = 16.77 ± 0.42 years), who were assigned to either a VR group (n = 11) or a control group (n = 11). The VR group completed 3 weekly sessions using the SensiballVR™ platform for a duration of 7 weeks, in addition to their regular training. Scanning frequency (before ball reception, during control, and off-the-ball) and passing performance (by execution type, outcome, and pass style) were assessed via video analysis in small-sided games pre- and post-intervention. Results: Between-group analysis revealed that the VR group achieved significantly greater improvements across all scanning domains compared to the control group, with mean-based percentage increases (calculated as the average of individual relative changes) ranging from + 198% to + 456%, rising from group mean values of 18 to 43 scans before ball reception, 20 to 43 during ball control, and 257 to 714 off the ball, versus − 12% to + 37% in controls (p < 0.01; ES = 0.58–0.83). Within-group analysis confirmed that scanning frequency increased significantly from pre- to post-test in the VR group (p = 0.003, ES = 0.89), whereas only off-the-ball scanning improved in the control group (p = 0.008, ES = 0.81). In passing performance, the VR group improved significantly in one-touch short-successful passes (+ 38%, p = 0.006, ES = 0.83), whereas the control group showed no meaningful change (+ 28%, p = 0.247, ES = 0.35). Although the VR group initially showed higher performance in control-pass short-successful passes (p = 0.038; ES = 0.44), this difference did not remain significant after false discovery rate correction (q = 0.199). Under the same condition, penetrative passes also improved significantly within the VR group (+ 108%, p = 0.029, ES = 0.66). No significant effects were observed for multi-touch passes (p > 0.05). Conclusion: Preliminary evidence suggests that immersive VR-based training can meaningfully enhance scanning behavior in youth football players, while potential benefits for passing outcomes remain exploratory. These findings highlight VR technology as a promising complementary tool in modern football development frameworks for improving perceptual-cognitive skills and decision-making.Clinical trial numberClinicalTrials.gov identifier: NCT07144371 (retrospectively registered on 27/08/2025).