Biology subjects

Marioni, R. E.

Publications and source records attributed to Marioni, R. E..

At least 19 recordsLinked to original sources

bioRxiv2026

Age-related dynamics of DNA co-methylation modules in humans

Age-related changes in the mean or variance of DNA methylation (DNAm) at individual CpG sites are well established. However, CpGs that share biological functions often exhibit coordinated regulation. Here, we analysed whole-blood DNAm data from over 7,500 individuals and identified modules of co-methylated CpGs that either maintain consistent correlations with age (age-stable modules) or undergo age-related remodelling (age-variable modules). We show that co-methylation modules are largely preserved during healthy aging, with only a small subset (~15%) showing substantial age-related reorganisation. Age-stable modules are enriched for CpG loci within promoters, transcription start sites, and CpG islands, whereas age-variable modules harbour loci that are depleted in these genomic domains and enriched for repressive chromatin marks. Age-stable co-methylation modules are also associated with long-range genetic regulation (trans-meQTLs), promoter-associated chromatin states, and higher-order chromatin interactions. While both age-stable and age-variable modules are enriched for developmental pathways, immune-related processes are distinctly associated with age-stable modules. Overall, our findings suggest that epigenetic aging is characterised by a spectrum of co-methylation network trajectories, ranging from stability to disruption, shaped by underlying genomic regulatory architecture.

systems biology↗

bioRxiv2026

Sex differences in the genetic basis of human recombination within 190,000 parent-child pairs

Structured AbstractO_ST_ABSIntroductionC_ST_ABSSexual reproduction uses a specialized cell division called meiosis, in which a single round of DNA replication is followed by two cell divisions to create hap-loid gametes. Genetic recombination in meiosis assures faithful segregation of chromosomes and establishes patterns of genetic linkage and inheritance. Meiotic recombination is thus a fundamental genomic process that shapes major features of the genomic landscape, influences mutation, and creates genetic diversity. RationaleThe genetic basis of between-individual variation in the occurrence and genomic location of meiotic recombination events remains poorly understood in humans. In particular, we lack an understanding of whether the same DNA regions shape vari-ation in recombination rates across studies. We address this by curating 112,144 maternal and 78,653 paternal meiosis events (parent-child pairs) across cohorts from four countries, creating the largest dataset of its kind. ResultsWe identify 58 independent genomic regions associated with the occurrence and genomic location of recombination events, 24 of which are unique across phe-notypes. Of the 24 unique regions, 6 are novel and specific to female meiotic recombination harboring genes linked to ovarian function, a 7th novel region is linked to male recombination. Many regions have significant sex-dependent ef-fects across studies. We estimate the between-sex genetic correlation for meiotic recombination rate to be 0.374 (0.072 SD), which captures the extent to which the genetic factors influencing a trait in one sex also impact the same trait in the other. 30-40% of the individual-level variation in recombination rate is at-tributable to DNA markers, of which only 37.7% (2.7% SD) and 47.1% (3.8% SD) is attributable to previously known global modulators of meiotic recombi-nation, in females and males respectively. We create a polygenic predictor that significantly predicts recombination rates of both women and men after adjusting for age and year of birth. We find significant correlations between female meiotic recombination, reproductive outcomes and reproductive aging, after adjusting for study ascertainment. ConclusionsIn the largest meta-analysis to date of human meiotic recombination, we identify novel loci with sex-specific effects that influence male and female recombination rate across multiple cohorts.

genomics↗

bioRxiv2025

Separating direct, indirect and parent-of-origin genetic effects in the human population

Here, we present a novel approach to estimate the degree to which the phenotypic effect of a DNA locus is attributable to four components: alleles in the child (direct genetic effects), alleles in the mother and the father (indirect genetic effects), or is dependent upon the parent from which it is inherited (parent-of-origin, PofO effects). Applying our model, JODIE, to 30,000 child-mother-father trios with phased DNA information from the Estonian Biobank (EstBB) and the Norwegian Mother, Father, Child Cohort (MoBa), we jointly estimate the phenotypic variance attributable to these four effects unbiased of assortative mating (AM) for height, body mass index (BMI) and childhood educational test score (EA). For all three traits, direct effects make the largest contribution to the genetic effect variance. But we find that parental indirect genetic effects make an equivalent combined contribution, and that there is a non-zero PofO effect variance for all traits. We calculate the heritability that would be obtained at the populationlevel in the absence of AM for common DNA loci, and show that the proportional contribution of direct effects to these heritability values can be calculated as 64.0% for EA in MoBa, 77.1% and 63.4% for height in MoBa and EstBB, and 81.2% and 88.0% for BMI in MoBa and EstBB. Additionally, using within-family genome-wide association testing, we identify 276 independently associated DNA regions that replicate across two additional biobanks, which all show a genotype-phenotype relationship that reflects an interplay of direct, indirect and PofO effects. Determining how direct, parental and PofO genetic effects combine across loci genome-wide to influence human phenotypic variation requires joint modeling of parental and child genotypes alongside the parental origin of loci and here, we make the first attempt to do this in the human population.

genetics↗

bioRxiv2024

MethylGPT: a foundation model for the DNA methylome

DNA methylation serves as a powerful biomarker for disease diagnosis and biological age assessment. However, current analytical approaches often rely on linear models that cannot capture the complex, context-dependent nature of methylation regulation. Here we present MethylGPT, a transformer-based foundation model trained on 226,555 (154,063 after QC and deduplication) human methylation profiles spanning diverse tissue types from 5,281 datasets, curated 49,156 CpG sites, and 7.6 billion training tokens. MethylGPT learns biologically meaningful representations of CpG sites, capturing both local genomic context and higher-order chromosomal features without external supervision. The model demonstrates robust methylation value prediction (Pearson R=0.929) and maintains stable performance in downstream tasks with up to 70% missing data. Applied to age prediction across multiple tissue types, MethylGPT achieves superior accuracy compared to existing methods. Analysis of the models attention patterns reveals distinct methylation signatures between young and old samples, with differential enrichment of developmental and aging-associated pathways. When finetuned to mortality and disease prediction across 60 major conditions using 18,859 samples from Generation Scotland, MethylGPT achieves robust predictive performance and enables systematic evaluation of intervention effects on disease risks, demonstrating potential for clinical applications. Our results demonstrate that transformer architectures can effectively model DNA methylation patterns while preserving biological interpretability, suggesting broad utility for epigenetic analysis and clinical applications.

systems biology↗

bioRxiv2024

Weighted 2D-kernel density estimations provide a new probabilistic measure for epigenetic age

BackgroundEpigenetic aging signatures can provide insights into the human aging process. Within the last decade many alternative epigenetic clocks have been described, which are typically based on linear regression analysis of DNA methylation at multiple CG dinucleotides (CpGs). However, this approach assumes that the epigenetic modifications follow either a continuous linear or logarithmic trajectory. In this study, we explored an alternative non-parametric approach using 2D-kernel density estimation (KDE) to determine epigenetic age. ResultsWe used Illumina BeadChip profiles of blood samples of various studies, exemplarily selected the 27 CpGs with highest linear correlation with chronological age (R2 > 0.7), and computed KDEs for each of them. The probability profiles for individual KDEs were further integrated by a genetic algorithm to assign an optimal weight to each CpG. Our weighted 2D-kernel density estimation model (WKDE) facilitated age-predictions with similar correlation and precision (R2 = 0.81, median absolute error = 4 years) as other commonly used clocks. Furthermore, our approach provided a variation score, which reflects the inherent variation of age-related epigenetic changes at different CpG sites within a given sample. An increase of the variation score by one unit reduced the mortality risk by 9.2% (95% CI (0.8387, 0.9872), P <0.0160) in the Lothian Birth Cohort 1921 after adjusting for chronological age and sex. ConclusionsWe describe a new method using weighted 2D-kernel density estimation (WKDE) for accurate epigenetic age-predictions and to calculate variation scores, which provide an additional variable to estimate biological age.

genomics↗

bioRxiv2023

A novel framework to build saliva-based DNA methylation biomarkers: quantifying systemic chronic inflammation as a case study

Accessible and non-invasive biomarkers that measure human ageing processes and the risk of developing age-related disease are paramount in preventative healthcare. In this study, we describe a novel framework to train saliva-based DNA methylation (DNAm) biomarkers that are reproducible and biologically interpretable. By leveraging a reliability dataset with replicates across tissues, we demonstrate that it is possible to transfer knowledge from blood DNAm data to saliva DNAm data using DNAm proxies of blood proteins (EpiScores). We then apply these methods to create a new saliva-based epigenetic clock (InflammAge) that quantifies systemic chronic inflammation (SCI) in humans. Using a large blood DNAm human cohort with linked electronic health records and over 18,000 individuals (Generation Scotland), we demonstrate that InflammAge significantly associates with all-cause mortality, disease outcomes, lifestyle factors and immunosenescence; in many cases outperforming the widely used SCI biomarker C-reactive protein (CRP). We propose that our biomarker discovery framework and InflammAge will be useful to improve our understanding of the molecular mechanisms underpinning human ageing and to assess the impact of gero-protective interventions.

bioinformatics↗

bioRxiv2023

Probabilistic inference of epigenetic age acceleration from cellular dynamics

The emergence of epigenetic predictors was a pivotal moment in geroscience, propelling the measurement and concept of biological ageing into a quantitative era. However, while current epigenetic clocks have shown strong predictive power, they do not reflect the underlying biological mechanisms driving methylation changes with age. Consequently, biological interpretation of their estimates is limited. Furthermore, our findings suggest that clocks trained on chronological age are confounded by non-age-related phenomena. To address these limitations, we developed a probabilistic model that describes methylation transitions at the cellular level. Our approach reveals two measurable components, acceleration and bias, that directly relate to perturbations of the underlying cellular dynamics. Acceleration is the proportional increase in the speed of methylation transitions across CpG sites, whereas bias is the degree of global change in methylation affecting all CpG sites uniformly. Using data from 7,028 participants from the Generation Scotland study, we found the age acceleration parameter to be associated with physiological traits known to impact healthy ageing. Furthermore, a genome-wide association study of age acceleration identified four genomic loci previously linked with ageing.

cell biology↗

bioRxiv2022

Refining epigenetic prediction of chronological and biological age

Epigenetic clocks can track both chronological age (cAge) and biological age (bAge). The latter is typically defined by physiological biomarkers and risk of adverse health outcomes, including all-cause mortality. As cohort sample sizes increase, estimates of cAge and bAge become more precise. Here, we aim to refine predictors and improve understanding of the epigenomic architecture of cAge and bAge. First, we perform large-scale (N = 18,413) epigenome-wide association studies (EWAS) of chronological age and all-cause mortality. Next, to improve cAge prediction, we use methylation data from 24,673 participants from the Generation Scotland (GS) study, the Lothian Birth Cohorts (LBC) of 1921 and 1936 and 8 publicly available datasets. Through the inclusion of linear and non-linear age-CpG associations from the EWAS, feature pre-selection/dimensionality reduction in advance of elastic net regression, and a leave-one-cohort-out (LOCO) cross validation framework, we arrive at an improved cAge predictor (median absolute error = 2.3 years across 10 cohorts). In addition, we train a predictor of bAge on 1,214 all-cause mortality events in GS, based on epigenetic surrogates for 109 plasma proteins and the 8 component parts of GrimAge, the current best epigenetic predictor of all-cause mortality. We test this predictor in four external cohorts (LBC1921, LBC1936, the Framingham Heart Study and the Womens Health Initiative study) where it outperforms GrimAge in its association to survival (HRGrimAge = 1.47 [1.40, 1.54] with p = 1.08 x 10-52, and HRbAge = 1.52 [1.44, 1.59] with p = 2.20 x 10-60). Finally, we introduce MethylBrowsR, an online tool to visualize epigenome-wide CpG-age associations.

genomics↗

bioRxiv2022

Epigenetic markers of adverse lifestyle identified among night shift workers

BackgroundEpigenetic changes in the form of DNA methylation (DNAm) may act as biological markers of risk factors or adverse health states. We investigated associations between night shift work and established DNAm predictors of lifestyle, and compared them with those observed between night shift work and self-reported or conventionally-measured phenotypes. MethodsIn two cohort studies, Generation Scotland (GS) (n=7,028) and Understanding Society (UKHLS) (n=1,175), we evaluated associations between night shift work and four lifestyle factors (body mass index, smoking, alcohol, education) using both conventionally-measured phenotypes and DNA methylation-based scores proxying the phenotypes. DNA methylation-based measures of biological ageing were also generated using six established "epigenetic clocks". Meta-analysis of GS and UKHLS results was conducted using inverse-variance weighted fixed effects. ResultsNight shift work was associated with higher BMI (0.79; 95%CI 0.02, 1.56; p=0.04) and lower education (-0.18; -0.30, -0.07; p=0.002). There was weak evidence of association between night shift work and DNAm scores for smoking (0.06, -0.03, 0.15; p=0.18) and education (-0.24; -0.49, 0.01; p=0.06) in fully adjusted models. Two of the epigenetic age measures demonstrated higher age acceleration among night shift workers (0.80; 0.42, 1.18; p<0.001 for GrimAge and 0.46; 0.00, 0.92; p=0.05 for PhenoAge). ConclusionsNight shift work is associated with phenotypic and DNAm-based measures of lower education. Night shift work was also related to DNAm predictors of smoking and ageing.

genetics↗

bioRxiv2022

The influence of biological and statistical properties of CpGs on epigenetic predictions of eighteen traits

BackgroundCpG methylation levels can help to explain inter-individual differences in phenotypic traits. Few studies have explored whether identifying CpG subsets based on biological and statistical properties can maximise predictions while minimising array content. MethodsVariance component analyses and penalised regression (epigenetic predictors) were used to test the influence of (i) the number of CpGs considered, (ii) mean CpG methylation variability and (iii) methylation QTL status on the variance captured in eighteen traits by blood DNA methylation. Training and test sets comprised [≤]4,450 and [≤]2,578 unrelated individuals from Generation Scotland, respectively. ResultsAs the number of CpG sites under consideration decreased, so too did the estimates from the variance components and prediction analyses. Methylation QTL status and mean CpG variability did not influence variance components. However, relative effect sizes were 15% larger for epigenetic predictors based on CpGs with methylation QTLs compared to sites without methylation QTLs. Relative effect sizes were 45% larger for predictors based on CpGs with mean beta-values between 10%-90% compared to those using hypo- or hypermethylated CpGs (beta-value [≤]10% or [≥]90%). ConclusionArrays with fewer CpGs could reduce costs, leading to increased sample sizes for analyses. Our results show that reducing array content can restrict prediction metrics and careful attention must be given to the biological and distribution properties of CpGs in array content selection.

genomics↗

bioRxiv2021

Longitudinal dynamics of clonal hematopoiesis identifies gene-specific fitness effects

The prevalence of clonal haematopoiesis of indeterminate potential (CHIP) in healthy individuals increases rapidly from age 60 onwards and has been associated with increased risk for malignancy, heart disease and ischemic stroke. CHIP is driven by somatic mutations in stem cells that are also drivers of myeloid malignancies. Since mutations in stem cells often drive leukaemia, we hypothesised that stem cell fitness substantially contributes to transformation from CHIP to leukaemia. Stem cell fitness is defined as the proliferative advantage over cells carrying no or only neutral mutations. It is currently unknown whether mutations in different CHIP genes lead to distinct fitness advantages that could form the basis for patient stratification. We set out to quantify the fitness effects of CHIP drivers over a 12 year timespan in older age, using longitudinal error-corrected sequencing data. We developed a new method based on drift-induced fluctuation (DIF) filtering to extract fitness effects from longitudinal data, and thus quantify the growth potential of variants within each individual. Our approach discriminates naturally drifting populations of cells and faster growing clones, while taking into account individual mutational context. We show that gene-specific fitness differences can outweigh inter-individual variation and therefore could form the basis for personalised clinical management.

cell biology↗

bioRxiv2020

DNA methylation proxies for 16 plasma proteins predict the incidence of 7 leading causes of morbidity

Protein biomarkers have been identified across many age-related morbidities. However, characterising epigenetic influences could further inform disease predictions. Here, we leverage epigenome-wide data to study links between the DNAm signatures of the circulating proteome and incident diseases. Using data from four cohorts, we trained and tested epigenetic scores (EpiScores) for 953 plasma proteins, identifying 109 scores that explained between 1% and 58% of the variance in protein levels after adjusting for known protein quantitative trait loci (pQTL) genetic effects. By projecting these EpiScores into an independent sample, (Generation Scotland; n=9,537) and relating them to incident morbidities over a follow-up of 14 years, we uncovered 137 EpiScore - disease associations. These associations were largely independent of immune cell proportions, common lifestyle and health factors and biological aging. Notably, we found that our diabetes-associated EpiScores highlighted previous top biomarker associations from proteome-wide assessments of diabetes. These EpiScores for protein levels can therefore be a valuable resource for disease prediction and risk stratification.

genomics↗

bioRxiv2020

A comparison of blood and brain-derived ageing and inflammation-related DNA methylation signatures and their association with microglial burdens

Inflammation and ageing-related DNA methylation patterns in the blood have been linked to a variety of morbidities, including cognitive decline and neurodegenerative disease. However, it is unclear how these blood-based patterns relate to patterns within the brain, and how each associates with central cellular profiles. In this study, we profiled DNA methylation in both the blood and in five post-mortem brain regions (BA17, BA20/21, BA24, BA46 and hippocampus) in 14 individuals from the Lothian Birth Cohort 1936. Microglial burdens were additionally quantified in the same brain regions. DNA methylation signatures of five epigenetic ageing biomarkers ( epigenetic clocks), and two inflammatory biomarkers (DNA methylation proxies for C-reactive protein and interleukin-6) were compared across tissues and regions. Divergent correlations between the inflammation and ageing signatures in the blood and brain were identified, depending on region assessed. Four out of the five assessed epigenetic age acceleration measures were found to be highest in the hippocampus ({beta} range=0.83-1.14, p[≤]0.02). The inflammation-related DNA methylation signatures showed no clear variation across brain regions. Reactive microglial burdens were found to be highest in the hippocampus ({beta}=1.32, p=5x10-4); however, the only association identified between the blood- and brain-based methylation signatures and microglia was a significant positive association with acceleration of one epigenetic clock (termed DNAm PhenoAge) averaged over all five brain regions ({beta}=0.40, p=0.002). This work highlights a potential vulnerability of the hippocampus to epigenetic ageing and provides preliminary evidence of a relationship between DNA methylation signatures in the brain and differences in microglial burdens.

neuroscience↗

bioRxiv2020

Variation in VKORC1 is associated with vascular dementia

Genetic variation in VKORC1 is associated with differences in the coagulation of blood and consequentially with sensitivity to the drug warfarin. Variation in VKORC1 has also been linked to parental dementia. However, it is unclear whether the relationship persists for the diagnosis in patients themselves, whether the association holds only for certain forms of dementia, and if those taking warfarin are at greater risk. Here, we use data from 211,423 participants from UK Biobank to examine the relationship between VKORC1, risk of dementia, and the interplay with warfarin use. We find that the T-allele in rs9923231 confers a greater risk for vascular dementia (OR=1.28, p=0.0069), but not for general dementia (OR=1.04, p=0.21) or Alzheimer dementia (OR=1.05, p=0.35), and that the risk of vascular dementia is not affected by warfarin use in carriers of the T-allele. Our study reports for the first time an association between rs9923231 and vascular dementia, but further research is warranted to explore potential mechanisms and specify the relationship between rs9923231 and features of vascular dementia.

genetics↗

bioRxiv2020

Rare Genetic Variants Underlie Outlying levels of DNA Methylation and Gene-Expression

Testing the effect of rare variants on phenotypic variation is difficult due to the need for extremely large cohorts to identify associated variants given expected effect sizes. An alternative approach is to investigate the effect of rare genetic variants on low-level genomic traits, such as gene expression or DNA methylation (DNAm), as effect sizes are expected to be larger for low-level compared to higher-order complex traits. Here, we investigate DNAm in healthy ageing populations - the Lothian Birth cohorts of 1921 and 1936 and identify both transient and stable outlying DNAm levels across the genome. We find an enrichment of rare genetic variants within 1kb of DNAm sites in individuals with stable outlying DNAm, implying genetic control of this extreme variation. Using a family-based cohort, the Brisbane Systems Genetics Study, we observed increased sharing of DNAm outliers among more closely related individuals, consistent with these outliers being driven by rare genetic variation. We demonstrated that outlying DNAm levels have a functional consequence on gene expression levels, with extreme levels of DNAm being associated with gene expression levels towards the tails of the population distribution. Overall, this study demonstrates the role of rare variants in the phenotypic variation of low-level genomic traits, and the effect of extreme levels of DNAm on gene expression.

genetics↗

bioRxiv2020

Integrative omics approach to identify the molecular architecture of inflammatory protein levels in healthy older adults

The molecular factors which control circulating levels of inflammatory proteins are not well understood. Furthermore, association studies between molecular probes and human traits are often performed by linear model-based methods which may fail to account for complex structure and interrelationships within molecular datasets. Therefore, in this study, we perform genome- and epigenome-wide association studies (GWAS/EWAS) on the levels of 70 plasma-derived inflammatory protein biomarkers in healthy older adults (Lothian Birth Cohort 1936; n = 876; Olink(R) inflammation panel). We employ a Bayesian framework (BayesR+) which can account for issues pertaining to data structure and unknown confounding variables (with sensitivity analyses using ordinary least squares- (OLS) and mixed model-based approaches). We identified 13 SNPs associated with 13 proteins (n = 1 SNP each) concordant across OLS and Bayesian methods. We identified three CpG sites spread across three proteins (n = 1 CpG each) that were concordant across OLS, mixed-model and Bayesian analyses. Tagged genetic variants accounted for up to 45% of variance in protein levels (for MCP2, 36% of variance alone attributable to one polymorphism). Methylation data accounted for up to 46% of variation in protein levels (for CXCL10). Up to 66% of variation in protein levels (for VEGFA) was explained using genetic and epigenetic data combined. We demonstrated putative causal relationships between CD6 and IL18R1 with inflammatory bowel disease, and between IL12B and Crohns disease. Our data may aid understanding of the molecular regulation of the circulating inflammatory proteome as well as causal relationships between inflammatory mediators and disease.

genomics↗

bioRxiv2020

Epigenetic clocks predict prevalence and incidence of leading causes of death and disease burden

Individuals of the same chronological age display different rates of biological ageing. A number of measures of biological age have been proposed which harness age-related changes in DNA methylation profiles. These include methylation-based predictors of chronological age (HorvathAge, HannumAge), all-cause mortality (DNAm PhenoAge, DNAm GrimAge) and telomere length (DNAm Telomere Length). In this study, we test the association between these epigenetic markers of ageing and the prevalence and incidence of the leading causes of disease burden and mortality in high-income countries. Furthermore, we test the clocks relationships with phenotypic measures associated with these conditions, including spirometric and biochemical traits. We carry out these analyses in 9,537 individuals from the Generation Scotland: Scottish Family Health Study. We find that DNAm GrimAge outperforms other epigenetic clocks in its associations with self-report disease prevalence and related clinical traits. DNAm GrimAge associates with chronic obstructive pulmonary disease (COPD) prevalence (Odds Ratio = 3.29, P = 3.0 x 10-4) and pulmonary spirometry tests ({beta} = [-0.10 to -0.15], P = [1.4 x 10-4 to 1.4 x 10-6]) at study baseline after adjusting for possibly confounding risk factors including alcohol, body mass index, deprivation, education and smoking. After adjusting for these confounding risk factors, DNAm GrimAge, DNAm PhenoAge and DNAm Telomere Length, measured at study baseline, predict incidence of ICD-10-coded disease states including COPD, type 2 diabetes and cardiovascular disease after thirteen years of follow-up (Hazard Ratios = [0.80 (telomere length) to 2.19 (GrimAge)], P = [9.9 x 10-4, 1.9 x 10-14]). Our data show that despite accounting for several possible confounding variables, epigenetic markers of ageing predict incidence of common disease. This may have significant implications for their potential utility in clinical settings to complement gold-standard methods of clinical assessment and management.

genomics↗

bioRxiv2019

DNA methylation in the APOE gene: its link with Alzheimer’s and cardiovascular health

Genetic variation in the apolipoprotein E (APOE) gene is associated with Alzheimers disease (AD) and risk factors for cardiovascular disease (CVD). DNA methylation at APOE has been linked to altered cognition and AD. It is unclear if epigenetic marks could be used for predicting future disease. We assessed blood-based DNA methylation at 13 CpGs in the APOE gene in 5828 participants from the Generation Scotland (GS) cohort. Using linear regression, we examined the relationship between APOE methylation, cognition, cholesterol, and the risks for AD and CVD. DNA methylation at two CpGs was associated with the ratio of total-to-HDL cholesterol, but not with cognition, or the risks of AD or CVD. APOE methylation could be involved in the levels of blood cholesterol, but there is no evidence for the utility of APOE methylation as a biomarker for predicting AD or CVD.

neuroscience↗