Robin Joaquim Hofmeister

Our understanding of the genetic architecture of obesity is primarily based on studies of adults, with sparser data from childhood. Here, we conducted age-stratified genetic association studies against objectively measured ( n = 62,276) or recalled childhood adiposity-related traits ( n _eff = 599,924), identifying 624 common variants associated with childhood adiposity, with one-third having no concordant association with adult body mass index. Signals linked to the leptin–melanocortin pathway ( BSX , GNAS , LEPR and PCSK1 ) and incretin signaling ( GIPR and GLP1R ) showed childhood-specific effects on adiposity. Single-nucleus RNA sequencing data identified childhood-specific adiposity-regulating cell populations in the arcuate nucleus and mammillary bodies, indicating neurocircuits that regulate adiposity specifically during childhood. Finally, sequencing in 479,615 individuals uncovered rare protein-coding variation in ADCY3 , CALCR , MC4R , MRAP2 , POMC and MYH13 , exhibiting stronger associations in childhood than in adults. Our findings suggest that childhood provides a more sensitive window for the study of key endocrine and neuropeptide pathways that regulate energy balance.

Assortative mating (AM), the tendency to choose partners with similar traits, contributes to the genetic architecture of human traits. Detecting its genetic footprint (genetic AM) requires either genotyped couples or detection of gametic phase disequilibrium (which is insensitive to recent changes). We introduce an inter-chromosomal phasing approach that reconstructs parental haplotypes from close relatives and correlates maternally and paternally inherited polygenic scores to provide estimates enriched for recent AM. Applied to 69 traits in 245,884 UK Biobank and 194,325 Estonian Biobank individuals, our approach showed concordance with mate-pair-based estimates comparable to or better than existing approaches, significantly outperforming them in the Estonian Biobank (through-origin r ² = 0.638 versus 0.479, P = 0.005) and performing comparably in the UK Biobank ( r ² = 0.737 versus 0.731). We identified genetic AM for traits such as height, education, overall health and time spent watching TV, with assortment for height and education accelerating in recent generations (for example, θ _HAP = 0.041 (0.034–0.048) versus 0.015 (0.007–0.024) for height in post-1980 versus pre-1960 Estonian Biobank cohorts).

Re-identification by phenotypic prediction aims to determine whether a genome belongs to a specific individual by comparing the individual’s known traits with those predicted from the genome. This type of tracing attack is widely discussed in the genomic privacy literature, yet previous studies have been criticized for overstating its practical risks. Over the past decade, genome-wide association studies (GWAS) with increasing sample size improved the accuracy of phenotypic prediction, potentially enhancing such attacks. To quantify their real-world threat, we developed a probabilistic framework that estimates the likelihood of a match between an individual’s observed traits and polygenic scores (PGS) derived from a genome, while accounting for prediction accuracy and genetic and environmental correlations between the traits. We benchmarked this re-identification method and examined how the prior probability (reflecting the a priori chance that a random genome and set of traits correspond to the same person) affects performance. Finally, we assessed whether sensitive information could be inferred through this attack by attempting to predict multiple sensitive haplotypes, such as APOE- ε 4 (linked with Alzheimer’s disease). Our re-identification method outperformed a state-of-the-art tool, and reached a precision above 99% for a recall of 40% when considering a prior of 50%. However, after considering real-world settings, we estimated that realistic priors would not exceed 4 × 10 ⁻⁴ %, resulting in a precision lower than 0.13% at the same recall (40%). The inference of sensitive genotypes also proved ineffective, as achieving a precision above 50% for identifying APOE- ε 4 carriers was only possible at a recall below 20%. To conclude, although re-identification by phenotypic prediction is technically feasible, our findings indicate that its effectiveness in real-world conditions is limited. These results counterpoint to earlier claims of severe genomic privacy risks and offer guidance for policymakers, biobank administrators, and research participants.

Assortative mating (AM), the tendency for individuals to choose partners with similar traits, plays an important role in social stratification and has a wide-spread impact on the genetic architecture of complex human traits. A genetic footprint of this behaviour, genetic assortative mating (GAM), has been documented for many traits. However, most existing approaches to estimate GAM either rely on genotyped couples – rare in large biobanks – or on estimates based on gametic phase disequilibrium (GPD), which reflect cumulative effects over multiple generations and cannot detect short-term events. We introduce a novel, scalable approach that infers genetic data for the parental generation to boost statistical power and improve interpretability when estimating GAM in biobank cohorts. Our method reconstructs the two parental haplotypes of biobank individuals by leveraging state-of-the-art inter-chromosomal phasing based on close relative information. By correlating polygenic scores computed separately on maternally and paternally inherited haplotypes, we infer the extent of GAM. Applied to 245,884 individuals from the UK Biobank and 194,325 individuals from the Estonian Biobank, our haplotype-based estimates showed strong concordance with GAM estimates from genotyped mate pairs across 69 traits and improved performance compared to existing GPD-based methods. We replicated previously identified patterns of GAM for many traits (educational attainment, height, BMI, and alcohol consumption), and revealed new ones (overall health and sedentary lifestyle). Temporal and geographic stratification revealed accelerated assortment in recent generations — particularly for education and height — and modest differences between urban and rural contexts. Our method enables scalable, interpretable, and generation-specific estimation of GAM in large biobank cohorts, providing new insights into the dynamic nature of mate choice and its impact on the genetic makeup of populations.

Copy-number variants (CNVs) are linked to a spectrum of outcomes and carriers of the same variant exhibit variable disease severity. We explored the impact of an individual’s polygenic score (PGS) on explaining these differences, focusing on 119 established CNV-trait associations involving 43 clinically-relevant phenotypes. We called CNVs among white British UK Biobank participants, then divided samples into a training set (n = 264,372) to derive independent PGS weights, and a CNV-carrier-enriched test set (n = 96,716). Assessing the individual, joint, and synergistic contribution of CNVs and PGSs, we identified a significant additive effect for 45 (38%) CNV-trait pairs, as well as two scale-dependent interactions between PGSs for gamma-glutamyltransferase levels and grip strength, and a 22q11.23 duplication encompassing multiple key genes in glutathione metabolism. A (spurious) negative correlation between an individual’s CNV carrier status and their PGS would be expected under selective participation-induced collider bias. Instead, we observed a widespread positive correlation, which could only be partially accounted for by linkage disequilibrium. Given a non-null inheritance rate for all 17 testable CNVs, we explored whether assortative mating could explain this positive CNV-PGS association. We found strong agreement between this correlation and the one predicted by assortment (r = 0.45, p = 2.0 × 10 ^-7 ). Similar suggestive trends of positive correlation were observed between PGSs and genome-wide burden of CNVs or rare loss-of-function variants. Our results demonstrate that PGSs contribute to the variable expressivity of CNVs and rare variants, and improve the identification of individuals at higher risk of clinically relevant comorbidities. We also highlight pervasive assortative mating as a likely mechanism contributing to the compounding of genetic effects across mutational classes.

Parent-of-origin effects (POEs) occur when the effect of a genetic variant depends on its parental origin¹. Traditionally linked to genomic imprinting, POEs are believed to occur due to parental conflict over resource allocation to offspring, resulting in opposing parental influences². Despite their importance, POEs remain underexplored in complex traits, owing to the lack of parental genomes. Here we present an approach to infer the parent of origin of alleles without parental genomes, leveraging interchromosomal phasing, mitochondrial and X chromosome data, and sex-specific crossover in siblings. Applied to the UK Biobank, this enabled parent-of-origin inference for up to 109,385 individuals. Genome-wide association study scans for 59 complex traits and over 14,000 protein quantitative trait loci contrasting maternal and paternal effects identified over 30 POEs and confirmed more than 50% of known associations. More than one third of these showed opposite parental influences, especially for traits related to growth (for example, IGF1 and height) and metabolism (for example, type 2 diabetes and triglyceride levels). Replication in up to 85,050 individuals from the Estonian Biobank and 42,346 offspring from the Norwegian Mother, Father and Child Cohort Study (MoBa) validated 87% of testable associations. Overall, our findings highlight the contribution of POEs to complex traits and support the parental conflict hypothesis, providing compelling evidence for this understudied evolutionary phenomenon.

Large biobanks with whole-genome sequencing (WGS) now enable the association of noncoding rare variants with complex human traits. Given that >98% of the genome is available for exploration, the selection of noncoding variants remains a critical yet unresolved challenge in these analyses. Here we leverage knowledge of blood gene regulation and deleteriousness scores to select noncoding variants pertinent for association with blood-related traits. Integrating WGS and 42 blood cell count and biomarker measurements for 166,740 UK Biobank samples, we perform variant collapsing tests, identifying hundreds of gene–trait associations involving noncoding variants. However, we demonstrate that most of these noncoding rare variant associations (1) reproduce associations known from previous studies and (2) are driven by linkage disequilibrium between nearby common and rare variants. This study underscores the prevailing challenges in rare variant analysis and the need for caution when interpreting noncoding rare variant association results.

Our understanding of the genetic architecture of obesity has primarily been shaped by observations from adult populations with relatively few studies on childhood obesity. To address this gap, we conduct a longitudinal genetic association study in up to ~600,000 individuals with objectively measured or recalled childhood adiposity-related traits. We identify 624 common variant signals (only 7% are previously reported) associated with childhood adiposity, of which one third have no concordant association with adult BMI. Signals linked to the leptin-melanocortin pathway (BSX, GNAS, LEPR and PCSK1) and incretin-signalling genes (GIPR and GLP1R) show stronger associations in childhood than in adults, suggesting that childhood provides a more sensitive window for detecting variation in key endocrine and neuropeptide pathways regulating energy balance. This observation is further supported by integrating single-nucleus RNA sequencing data from the human hypothalamus, identifying childhood-specific adiposity-regulating cell populations in the arcuate nucleus and mammillary bodies, indicating distinct neuro-circuits that regulate adiposity only during childhood. Three signals showed parent-of-origin specific associations with childhood BMI, at KLF14 (maternal-specific), GNAS (parental-discordant) and ZDBF2 (paternal-specific). Finally, we complement these common variant analyses with DNA sequence data in 479,615 individuals, identifying rare protein-coding variation in ADCY3, CALCR, MC4R, MRAP2, POMC and MYH13, all of which demonstrate stronger adiposity associations in childhood than in adults. Collectively, our findings emphasize the value of expanding research on childhood adiposity alongside studies focused on adult obesity measures.

Our understanding of the genetic architecture of obesity has primarily been shaped by observations from adult populations with relatively few studies on childhood obesity. To address this gap, we conduct a longitudinal genetic association study in up to ∼600,000 individuals with objectively measured or recalled childhood adiposity-related traits. We identify 624 common variant signals (only 7% are previously reported) associated with childhood adiposity, of which one third have no concordant association with adult BMI. Signals linked to the leptin-melanocortin pathway ( BSX, GNAS , LEPR and PCSK1 ) and incretin-signalling genes ( GIPR and GLP1R ) show stronger associations in childhood than in adults, suggesting that childhood provides a more sensitive window for detecting variation in key endocrine and neuropeptide pathways regulating energy balance. This observation is further supported by integrating single-nucleus RNA sequencing data from the human hypothalamus, identifying childhood-specific adiposity-regulating cell populations in the arcuate nucleus and mammillary bodies, indicating distinct neuro-circuits that regulate adiposity only during childhood. Three signals showed parent-of-origin specific associations with childhood BMI, at KLF14 (maternal-specific), GNAS (parental-discordant) and ZDBF2 (paternal-specific). Finally, we complement these common variant analyses with DNA sequence data in 479,615 individuals, identifying rare protein-coding variation in ADCY3 , CALCR , MC4R , MRAP2 , POMC and MYH13, all of which demonstrate stronger adiposity associations in childhood than in adults. Collectively, our findings emphasize the value of expanding research on childhood adiposity alongside studies focused on adult obesity measures.

Large biobanks have set a new standard for research and innovation in human genomics and implementation of personalized medicine. The Estonian Biobank was founded a quarter of a century ago, and its biological specimens, clinical, health, omics, and lifestyle data have been included in over 800 publications to date. What makes the biobank unique internationally is its translational focus, with active efforts to conduct clinical studies based on genetic findings, and to explore the effects of return of results on participants. In this review, we provide an overview of the Estonian Biobank, highlight its strengths for studying the effects of genetic variation and quantitative phenotypes on health-related traits, development of methods and frameworks for bringing genomics into the clinic, and its role as a driving force for implementing personalized medicine on a national level and beyond.