Population stratification is one of the source of inflation in epigenome-wide association studies (EWAS) when not properly accounted for. To address this, we developed methylation population scores (MPSs) to predict genetic principal components (GPCs) using a feature selection approach. We used multi-ethnic DNA methylation data from Illumina EPIC arrays across five cohorts, including MESA (n = 929), CARDIA (n = 1123), JHS (n = 1365), ARIC (n = 2338), and HCHS/SOL (n = 1475), randomly splitting participants into training (85%) and test (15%) sets. Within each cohort, associations between GPCs and CpG sites were estimated using linear regression adjusting for age, sex, smoking and alcohol use, race/ethnicity, body mass index, and cell type proportions, followed by meta-analysis and selection of CpGs with FDR <0.05. We then applied a two-stage weighted least squares Lasso regression to construct MPSs, adjusting for the aforementioned covariates. In the test dataset, MPSs showed strong correlation with GPCs, with R² ranging from 0.27 (MPS7 vs. GPC7) to 0.98 (MPS1 vs. GPC1). Visualization demonstrated that MPSs recapitulated the pattern shown by GPCs in differentiating self-reported White, Black, and Hispanic/Latino groups and outperformed methylation-based principal components constructed using alternative published methods. Additionally, MPSs showed comparable performance to GPCs in reducing inflation in EWAS. Overall, MPSs uses supervised learning with covariate adjustment to capture genetic structure across diverse populations, and provide a reliable estimate of population structure in the data and can complement GPCs when genetic data are absent.

Polygenic risk score values vary with genetic ancestry due to differences in population-specific allele frequencies and linkage disequilibrium patterns. We present a framework to calibrate polygenic risk scores based on ancestral makeup. We propose the "expected polygenic risk score" or ePRS, defined as the expected value of a polygenic risk score based on one's global or local admixture patterns. We further define the "residual polygenic risk score" or rPRS as measuring the deviation of the polygenic risk score from the ePRS. The ePRS reflects the baseline ancestry-driven component of genetic risk, whereas the rPRS isolates an ancestry-agnostic measure of genetic liability. Simulation studies confirm that it suffices to adjust for ePRS to obtain nearly unbiased estimates of the polygenic risk score-outcome association without further adjusting for principal components. Using the TOPMed and the All of Us datasets, effect size estimates for the rPRS (adjusted for ePRS) are similar to those obtained from polygenic risk scores adjusting for genetic principal components. The ePRS framework can protect from population stratification in association analysis and provide an equitable strategy to interpret genetic risk across diverse populations.

In admixed populations, formed by a mixing of two or more previously isolated populations, genomic segments can be traced to their ancestral populations ('ancestries'). Admixture mapping (AM) associates local ancestry with outcomes in admixed populations, detecting signals when causal variants differ in frequency or effect across ancestral populations. Prior work showed that adjusting for nearby GWAS-identified common variants does not fully explain some AM signals. Here, we assessed two approaches to explain the previously unexplained AM signal: (1) including sets of rare variants; (2) increasing the genomic region considered when searching for common variants. We studied these hypotheses comprehensively using a whole-genome sequencing dataset coupled with metabolomics from the Hispanic Community Health Study/Study of Latinos. From 16 genomic regions of interest, we identified 35 rare-variant sets associated with metabolite levels, of which 4 associations replicated. The locations of identified rare variant sets are either within the same gene or in proximity to adjusted common variants. Overall, our study provides new insight into the genetic architecture underlying AM signals via evaluating the contributions of common and rare variant sets. By integrating rare-variant analysis within the AM framework, our work highlights the potential interplay between variant types and establishes a novel strategy to guide future fine-mapping and causal variant identification in admixed populations. Yet the detected rare variants appear to explain only a small fraction of the AM signal, while inclusion of common variants from a larger genomic region appears to explain most of the AM signals.

BACKGROUND: The APOE ε4 allele has weaker association with ADRD outcomes in Hispanic/Latino populations compared to non-Hispanic White and East Asian populations.

METHODS: We used whole-genome sequencing data from 13,025 Hispanic/Latino adults to infer population groups (clusters) from short DNA sequences at the APOE gene region. We estimated the associations between these clusters and mild cognitive impairment (MCI) and age-adjusted significant cognitive decline assessed via the Study of Latinos-Investigation of Neurocognitive Aging (SOL-INCA) study (n = 5334). We compared these associations to those of the conventional APOE alleles.

RESULTS: We identified three clusters of individuals. Relative to the largest Cluster 1, being in Cluster 3 had higher likelihood of having significant cognitive decline, with odds ratio (OR) = 1.47, 95% confidence interval (CI) [1.18, 1.84], p < 0.001. The association with having prevalent MCI at the SOL-INCA exam was not statistically significant, with OR = 1.33, 95% CI [0.97, 1.83], p = 0.08. Associations of the ε4 allele with these traits were not statistically significant.

DISCUSSION: The new APOE profile has stronger risk association with age-adjusted significant cognitive decline in SOL-INCA Hispanics/Latinos compared to the ε4 allele.

Integrative approaches that simultaneously model multi-omics data have gained increasing popularity because they provide holistic system biology views of multiple or all components in a biological system of interest. Canonical correlation analysis (CCA) is a correlation-based integrative method designed to extract latent features shared between multiple assays by finding the linear combinations of features-referred to as canonical variables (CVs)-within each assay that achieve maximal across-assay correlation. Although widely acknowledged as a powerful approach for multi-omics data, CCA has not been systematically applied to multi-omics data in large cohort studies, which has only recently become available. Here, we adapted sparse multiple CCA (SMCCA), a widely-used derivative of CCA, to proteomics and methylomics data from the Multi-Ethnic Study of Atherosclerosis (MESA) and Jackson Heart Study (JHS). To tackle challenges encountered when applying SMCCA to MESA and JHS, our adaptations include the incorporation of the Gram-Schmidt (GS) algorithm with SMCCA to improve orthogonality among CVs, and the development of Sparse Supervised Multiple CCA (SSMCCA) to allow supervised integration analysis for more than two assays. Effective application of SMCCA to the two real datasets reveals important findings. Applying our SMCCA-GS to MESA and JHS, we identified strong associations between blood cell counts and protein abundance, suggesting that adjustment of blood cell composition should be considered in protein-based association studies. Importantly, CVs obtained from two independent cohorts also demonstrate transferability across the cohorts. For example, proteomic CVs learned from JHS, when transferred to MESA, explain similar amounts of blood cell count phenotypic variance in MESA, explaining 39.0%   50.0% variation in JHS and 38.9%   49.1% in MESA. Similar transferability was observed for other omics-CV-trait pairs. This suggests that biologically meaningful and cohort-agnostic variation is captured by CVs. We anticipate that applying our SMCCA-GS and SSMCCA on various cohorts would help identify cohort-agnostic biologically meaningful relationships between multi-omics data and phenotypic traits.

Allele frequency estimates in admixed populations, such as Hispanics and Latinos, rely on the sample's specific admixture composition and thus may differ between two seemingly similar populations. However, ancestry-specific allele frequencies, i.e., pertaining to the ancestral populations of an admixed group, may be particularly useful for prioritizing genetic variants for genetic discovery and personalized genomic health. We developed a method, ancestry-specific allele frequency estimation in admixed populations (AFA), to estimate the frequencies of biallelic variants in admixed populations with an unlimited number of ancestries. AFA uses maximum-likelihood estimation by modeling the conditional probability of having an allele given proportions of genetic ancestries. It can be applied using either local ancestry interval proportions encompassing the variant (local-ancestry-specific allele frequency estimations in admixed populations [LAFAs]) or global proportions of genetic ancestries (global-ancestry-specific allele frequency estimations in admixed populations [GAFAs]), which are easier to compute and are more widely available. Simulations and comparisons to existing software demonstrated the high accuracy of the method. We implemented AFA on high-quality imputed data of ∼9,000 Hispanics and Latinos from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), an understudied, admixed population with three predominant continental ancestries: Amerindian, European, and African. Comparison of the European and African estimated frequencies to the respective gnomAD frequencies demonstrated high correlations (Pearson R2 = 0.97-0.99). We provide a genome-wide dataset of the estimated ancestry-specific allele frequencies for available variants with allele frequency between 5% and 95% in at least one of the three ancestral populations. Association analysis of Amerindian-enriched variants with cardiometabolic traits identified five loci associated with lipid traits in Hispanics and Latinos, demonstrating the utility of ancestry-specific allele frequencies in admixed populations.

In a multi-stage analysis of 52,436 individuals aged 17-90 across diverse cohorts and biobanks, we train, test, and evaluate a polygenic risk score (PRS) for hypertension risk and progression. The PRS is trained using genome-wide association studies (GWAS) for systolic, diastolic blood pressure, and hypertension, respectively. For each trait, PRS is selected by optimizing the coefficient of variation (CV) across estimated effect sizes from multiple potential PRS using the same GWAS, after which the 3 trait-specific PRSs are combined via an unweighted sum called "PRSsum", forming the HTN-PRS. The HTN-PRS is associated with both prevalent and incident hypertension at 4-6 years of follow up. This association is further confirmed in age-stratified analysis. In an independent biobank of 40,201 individuals, the HTN-PRS is confirmed to be predictive of increased risk for coronary artery disease, ischemic stroke, type 2 diabetes, and chronic kidney disease.

Polygenic risk scores (PRS) are commonly used to quantify the inherited susceptibility for a trait, yet they fail to account for non-linear and interaction effects between single nucleotide polymorphisms (SNPs). We address this via a machine learning approach, validated in nine complex phenotypes in a multi-ancestry population. We use an ensemble method of SNP selection followed by gradient boosted trees (XGBoost) to allow for non-linearities and interaction effects. We compare our results to the standard, linear PRS model developed using PRSice, LDpred2, and lassosum2. Combining a PRS as a feature in an XGBoost model results in a relative increase in the percentage variance explained compared to the standard linear PRS model by 22% for height, 27% for HDL cholesterol, 43% for body mass index, 50% for sleep duration, 58% for systolic blood pressure, 64% for total cholesterol, 66% for triglycerides, 77% for LDL cholesterol, and 100% for diastolic blood pressure. Multi-ancestry trained models perform similarly to specific racial/ethnic group trained models and are consistently superior to the standard linear PRS models. This work demonstrates an effective method to account for non-linearities and interaction effects in genetics-based prediction models.

Polygenic risk scores (PRSs) are weighted sums of risk allele counts of single-nucleotide polymorphisms (SNPs) associated with a disease or trait. PRSs are typically constructed based on published results from Genome-Wide Association Studies (GWASs), and the majority of which has been performed in large populations of European ancestry (EA) individuals. Although many genotype-trait associations have generalized across populations, the optimal choice of SNPs and weights for PRSs may differ between populations due to different linkage disequilibrium (LD) and allele frequency patterns. We compare various approaches for PRS construction, using GWAS results from both large EA studies and a smaller study in Hispanics/Latinos: The Hispanic Community Health Study/Study of Latinos (HCHS/SOL, n = 12 , 803 ). We consider multiple approaches for selecting SNPs and for computing SNP weights. We study the performance of the resulting PRSs in an independent study of Hispanics/Latinos from the Women's Health Initiative (WHI, n = 3 , 582 ). We support our investigation with simulation studies of potential genetic architectures in a single locus. We observed that selecting variants based on EA GWASs generally performs well, except for blood pressure trait. However, the use of EA GWASs for weight estimation was suboptimal. Using non-EA GWAS results to estimate weights improved results.

Linear mixed models (LMMs) are widely used in genome-wide association studies (GWASs) to account for population structure and relatedness, for both continuous and binary traits. Motivated by the failure of LMMs to control type I errors in a GWAS of asthma, a binary trait, we show that LMMs are generally inappropriate for analyzing binary traits when population stratification leads to violation of the LMM's constant-residual variance assumption. To overcome this problem, we develop a computationally efficient logistic mixed model approach for genome-wide analysis of binary traits, the generalized linear mixed model association test (GMMAT). This approach fits a logistic mixed model once per GWAS and performs score tests under the null hypothesis of no association between a binary trait and individual genetic variants. We show in simulation studies and real data analysis that GMMAT effectively controls for population structure and relatedness when analyzing binary traits in a wide variety of study designs.