Alicia R. Martin

Environmental exposures shape disease risk, but the generalizability of circulating protein-based readouts of exposures across diverse populations remains unknown. Using plasma proteomic profiles of 5,416 proteins in 10,043 participants from the All of Us Research Program, we evaluated associations with and composite scores of 100+ exposures, spanning adiposity, air pollution, ambient temperature, and substance use. We find that proteomic exposure signatures are largely consistent across populations and socioeconomic strata, with 2.5 protein scores captured within-region exposure variation and outperformed geographic and socioeconomic correlates, but spatial confounders remain embedded within protein-environment associations. Scores trained in UK Biobank transferred to All of Us for biologically proximal exposures, supporting cross-biobank portability. These findings establish plasma proteomics as a scalable framework for exposure measurement across diverse populations, while underscoring that responsible applications require careful consideration of biological context, social patterning, and cohort structure.

Advances in high-throughput proteomics technologies have enabled the assessment of dynamic health states across biobank-scale cohorts. Disease prediction models built on these data have higher accuracy than baseline clinical models for a broad range of diseases, and provide avenues to understand disease pathogenesis. However, the generalizability of these prediction models across cohorts has yet to be established at scale. Here, we train models for 15 diseases in the UK Biobank (UKB; n=53,026) based on Olink proteomics data, and find high accuracy for disease prediction (mean AUC=0.74; range 0.56-0.89). These models have improved accuracy over models using only clinical factors (mean ΔAUC = 0.03), and do not depend on model architecture, with simpler models (e.g. L2) performing as well as complex models (e.g. transformer). We assess generalizability of the UKB-trained models in two external cohorts: FinnGen (FG; n=5,865) and All of Us (AoU; n=7,405), spanning two Olink platforms. We find that proteomics-based prevalent (classification) and incident (prediction over next 5 years) disease models largely generalize across cohorts, but performance varies across diseases. Specifically, 10 of 14 prevalent and 13 of 15 incident disease models show no significant decrease in performance across any biobank. Adjusting for demographic, ancestry, and technical covariates, we demonstrate that differences in phenotyping quality are likely the major drivers of variability across cohorts. These results indicate that proteomic risk models can be powerful and generalizable predictors of disease across multiple cohorts.

Genetic studies of serious mental illness (SMI, including schizophrenia and bipolar disorder) have implicated two classes of risk variant: common variants ^1–3 , each of small effect but collectively accounting for most population-level risk, and ultra-rare coding variants ^4,5 which confer large risk in a small number of individuals. Most such studies cannot directly assess the impact of specific variants on SMI, transdiagnostically, as the phenotypic information that they collect is focused on a single diagnosis. Here we report results of association analysis of predicted deleterious variants from exome sequencing of 21,958 cases (SMI and severe major depressive disorder) and 19,826 controls from the Mision Origen biobank in the Paisa genetic isolate of Colombia. We identified transdiagnostic associations to seven genes, three at an exome-wide and four at a false discovery rate significance threshold. Five of these genes have not previously been implicated in SMI ( ADAP1 , WIPI1 , NFE2L1 , FBXO10 , and HLA-DPB1 ); for the first four of these the association is driven by variants enriched in the Paisa due to a founder effect, filling the gap between ultra-rare and common variants. The strongest association that we observe is to a frameshift variant (7:904169:T:TC) in ADAP1 that is genome-wide significant for SMI, schizophrenia, and bipolar disorder and increases risk for both diagnoses by more than threefold; it truncates ADAP1, a brain-enriched Arf6 GTPase-activating/scaffolding protein that has been mechanistically linked to post-mitotic neuronal morphogenesis and regulation of dendritic differentiation ⁶ . Variant carriers differ significantly from other SMI cases in Mision Origen for psychotic, manic, and cognitive item-level phenotypes assessed through analyses of electronic health records ^7–9 and displayed impaired executive function on neurocognitive testing. Variant carriers differ significantly from other SMI cases in Mision Origen for multiple item-level phenotypes (psychotic, manic, and cognitive) assigned through natural language processing of longitudinal electronic health records. Although the 7:904169:T:TC variant is not significantly associated with major depressive disorder, carriers compared to others with this diagnosis display a significant increase in symptoms typical of bipolar disorder. The variant’s high allele frequency in Mision Origen (∼0.4%) will enable population-level studies to stratify SMI based on causation and identify additional factors influencing risk and resilience.

Major depressive disorder (MDD) significantly contributes to the mental health burden in Africa, requiring insights into prevalence and risk factors to inform culturally responsive care. This study assessed current and lifetime MDD prevalence and associated factors among 7,073 adults from hospital settings in Uganda, Kenya, Ethiopia, and South Africa within the NeuroGAP-Psychosis study. Prevalence was 0.9% for current MDD and 5.1% for lifetime MDD, substantially lower than prior estimates from the region. Lower rates may be due to the population studied or cultural variations in expression of symptoms. Multilevel logistic regression identified significant associations between current MDD and negative life events, alcohol use, and Kessler psychological distress scores, while lifetime MDD was linked to age, female sex, chronic pain, frequent headaches, and positive psychosis screening. These findings underscore the need for targeted mental health interventions tailored to the identified factors to reduce the burden of MDD in diverse African contexts.

Admixed individuals have been understudied in medical research largely due to their complex genetic ancestries. However, the consideration of admixture can identify ancestry-enriched genetic associations, delineating genetic underpinnings of cross-population phenotypic variation. Here, we performed admixture mapping in individuals with inferred admixture from African and European populations (N = 48,921). Across 22 traits, we identified 71 ancestry-trait associations, including loci where ancestral haplotypes explained phenotypic variation yet were missed by single-variant association testing due to their stricter multiple testing burden. One such locus where inferred local AFR ancestries are associated with increased hemoglobin A1c (HbA1c) was 12q14.3, highlighting its potential role in explaining differences between populations. Together, our results expand upon the phenotypic differences between populations and characterize loci where genetic ancestries play a critical role in the architecture of disease.

Biobanks increasingly include individuals with admixed genomes, yet conventional genome-wide association study frameworks either exclude participants who cannot be confidently assigned to a discrete ancestry group or ignore ancestry-specific effects. We present FELIX, a scalable framework for local-ancestry-aware genetic analysis that retains all participants without requiring discrete ancestry assignment. FELIX combines a compact ancestry-resolved genotype representation (FELIXla) with an adaptive association test that jointly evaluates shared-effect and ancestry-specific models at each variant (FELIXassoc). Simulations demonstrated well-calibrated inference under case-control imbalance and power that adapted to the locus-optimal model. Across 24 phenotypes in 240,038 All of Us participants, FELIX analyzed the 12.1% of individuals excluded by global-ancestry clustering and identified 15.4% more genome-wide significant loci than global-ancestry meta-analysis. Additional discoveries arose from recovering ancestry-specific haplotypes carried by admixed participants and from detecting ancestry-dependent marginal effects. Full-cohort effect estimates also improved polygenic score prediction across ancestries and traits.

Thyroid cancer is the most common endocrine malignancy, yet its biological underpinnings remain incompletely understood. Here we show that common risk alleles for thyroid cancer point to distinct biological pathways underlying disease susceptibility. We perform a multi-ancestry genome-wide association meta-analysis of thyroid cancer (16,167 cases and 2,430,374 controls), identifying 51 independent loci, including 21 not previously reported. By integrating these loci with genetic associations for 151 thyroid-cancer-related traits, we identify pleiotropic mechanistic clusters linked to thyroid function, oncogenic pathways, and mixed physiological function. Two thyroid-specific clusters, associated with thyroid stimulating hormone or thyroid growth and function, are enriched in thyroid tissues. Oncogenic clusters include DNA repair ( ATM , CHEK2 , TP53 ) and telomere maintenance ( TERT ) genes, implicating shared cancer mechanisms. Cluster-specific polygenic scores are associated with thyroid disease, cancer, and metabolic traits across ancestry groups, suggesting distinct genetic subtypes of thyroid cancer risk and supporting pleiotropy-based approaches to genetic risk stratification.

Here we developed and deployed the blended genome exome (BGE) method, a DNA library approach that generates low-pass whole-genome (1–4× mean depth) and deep whole-exome (30–40× mean depth) data in a single sequencing run. BGE is cost-effective, empowers most genomic discoveries possible with deep whole-genome sequencing and captures global common single-nucleotide polymorphism diversity. We applied BGE to sequence >53,000 samples from the PUMAS Project (Populations Underrepresented in Mental Illness Associations Studies), including African, African American and Latin American populations. Imputed genotypes showed high concordance with Illumina Global Screening Array calls ( R ² ≥ 95% for minor allele frequency ≥1%; ≥90% for minor allele frequency <1%), with consistent performance across local ancestries in admixed cohorts. For protein-coding copy number variants, deletions and duplications spanning at least three exons had a positive predicted value of ~90% relative to deep whole-genome data. At ~28% of the cost of deep whole-genome sequencing, BGE provides a scalable, reliable platform to expand genomic discovery and equitable access to sequencing in underrepresented populations.

Polygenic scores (PGS) are typically derived from single-trait genome-wide association studies (GWAS), yet many complex diseases arise from shared genetic liability distributed across correlated clinical dimensions. Accordingly, disease risk depends not only on how genetic liability is represented but also on the social context in which that liability is expressed. Whether phenome-derived latent factors improve prediction, and how social determinants of health (SDoH) modify the realized utility of PGS, remains unclear. Here we constructed PGS for 35 orthogonal latent phenomic factors derived from 2,772 phenotypes in 361,114 UK Biobank (UKB) participants and evaluated their phenomic specificity, cross-dataset portability and predictive performance relative to conventional disease-specific PGS across the UKB holdout, Mass General Brigham Biobank and the All of Us (AoU) Research Program. Factor-based PGS showed widespread, biologically coherent phenome-wide associations that were reproducible across biobanks and ancestries. Their predictive utility, however, was strongly disease dependent. For asthma, a respiratory factor PGS outperformed an internally derived disease-specific PGS and showed superior cross-ancestry portability, retaining 41.5% of European-ancestry predictive accuracy in African-ancestry individuals, compared with 22.9% for an asthma PGS derived from the largest available multi-ancestry GWAS. By contrast, disease-specific PGS remained superior for coronary artery disease (CAD) and type 2 diabetes (T2D). These findings suggest that phenome-derived aggregation is most beneficial when disease-specific GWAS incompletely capture underlying liability, including settings of biological heterogeneity or imprecise phenotyping. We then evaluated SDoH in AoU as a complementary axis shaping prevalent disease prediction beyond genetic susceptibility. Across all three diseases, SDoH contributed substantial and largely independent predictive information beyond the disease-optimal genetic model. SDoH also modified how genetic liability translated into observed disease prevalence: for asthma and CAD, genetic stratification attenuated with increasing social burden, whereas this attenuation was substantially weaker for T2D. As a result, the same genetic percentile corresponded to different standardized predicted prevalences across social strata, reflecting disease-specific shifts in baseline prevalence, genetic gradients and calibration. Together, these findings indicate that disease risk is shaped by both genetic liability and the social context in which that liability is realized. Phenome-derived PGS improve prediction under specific architectural conditions, whereas social context independently modifies the performance, calibration and interpretation of genetic risk across populations.