Biology subjects

Jarvelin, M.-R.

Publications and source records attributed to Jarvelin, M.-R..

At least 19 recordsLinked to original sources

bioRxiv2022

Genetic fine-mapping from summary data using a non-local prior improves detection of multiple causal variants

Genome-wide association studies (GWAS) have been successful in identifying genomic loci associated with complex traits. Genetic fine-mapping aims to detect independent causal variants from the GWAS-identified loci, adjusting for linkage disequilibrium (LD) patterns. The use of GWAS summary statistics and an adequate LD reference enable large sample sizes for fine-mapping, without direct access to individual-level data. We present FiniMOM (fine-mapping using a product inverse-moment priors), a novel Bayesian fine-mapping method for summarized genetic associations. For causal effects, the method uses a non-local inverse-moment prior, which is a natural prior distribution to model non-null effects in finite samples. A beta-binomial prior is set for the number of causal variants, with a parameterization that can be used to control for potential misspecifications in the LD reference. We test the performance of our method against a current state-of-the-art fine-mapping method SuSiE (sum-of-single-effects) across a range of simulated scenarios aimed to mimic a typical GWAS on circulating protein levels, and an applied example. The results show improved credible set coverage and power of the proposed method, especially in the case of multiple causal variants within a locus. The superior performance and the flexible parameterization to control for misspecified LD reference make FiniMOM a competitive alternative to other fine-mapping methods for summarized genetic data.

bioinformatics↗

bioRxiv2021

Adolescents' dietary habits and meal patterns influence school performance in the Northern Finland Birth Cohort 1986: mendelian randomisation study

Several observational studies indicate that dietary habits in children and adolescents are associated with school performance. These associations are heavily confounded by socio-economic characteristics, such as household income and parents educational attainment, amongst other factors. In this study, we report observational and causal effects of habitual diet on school performance, using individual level data for 9,220 adolescents in the Northern Finland Birth Cohort 1986. For this purpose, we derived principal components for the dietary variables, meal patterns and school performance variables. The observational study showed a significant association of consumption of foods that are high in fat, salt and sugar (HFSS) with poor performance in all school subjects, and an association of consumption of healthy foods and traditional foods with good school performance in general subjects, science and physical education (PE). Moreover, a positive association was observed between not skipping breakfast and good performance in all school subjects. Mendelian randomisation analysis confirmed a negative effect of HFSS on school performance in general/science subjects (-0.080, -0.128 to -0.033) and a positive effect of healthy food on school performance in general/science subjects (0.071, 0.024 to 0.119) and PE (0.065, 0.021 to 0.110). To conclude, we identified compelling evidence that HFSS foods and healthy foods were causally affecting school performance.

bioinformatics↗

bioRxiv2020

Signatures Of TSPAN8 Variants Associated With Human Metabolic Regulation And Diseases

Here, with the example of common copy number variation (CNV) in the TSPAN8 gene, we present an important piece of work in the field of CNV detection, CNV association with complex human traits such as 1H NMR metabolomic phenotypes and an example of functional characterization of CNVs among human induced pluripotent stem cells (HipSci). We report TSPAN8 exon 11 as a new locus associated with metabolomic regulation and show that its biology is associated with several metabolic diseases such as type 2 diabetes (T2D), obesity and cancer. Our results further demonstrate the power of multivariate association models over univariate methods and define new metabolomic signatures for several new genomic loci, which can act as a catalyst for new diagnostics and therapeutic approaches.

genomics↗

bioRxiv2020

Common maternal and fetal genetic variants show expected polygenic effects on the probability of being born small- or large-for-gestational-age (SGA or LGA), except in the smallest 3% of babies

Babies born clinically Small- or Large-for-Gestational-Age (SGA or LGA; sex- and gestational age-adjusted birth weight (BW) <10th or >90th percentile, respectively), are at higher risks of complications. SGA and LGA include babies who have experienced growth-restriction or overgrowth, respectively, and babies who are naturally small or large. However, the relative proportions within each group are unclear. We aimed to assess the extent to which the genetics of normal variation in birth weight influence the probability of SGA/LGA. We calculated independent fetal and maternal genetic scores (GS) for BW in 12,125 babies and 5,187 mothers. These scores capture the direct fetal and indirect maternal (via intrauterine environment) genetic contributions to BW, respectively. We also calculated maternal fasting glucose (FG) and systolic blood pressure (SBP) GS. We tested associations between each GS and probability of SGA or LGA. For the BW GS, we used simulations to assess evidence of deviation from an expected polygenic model. Higher BW GS were strongly associated with lower odds of SGA and higher odds of LGA (ORfetal=0.65 (0.60,0.71) and 1.47 (1.36,1.59); ORmaternal=0.80 (0.76,0.87) and 1.23 (1.15,1.31), respectively per 1 decile higher GS). Associations were in accordance with a polygenic model except in the smallest 3% of babies (Pfetal=0.0034, Pmaternal=0.023). Higher maternal GS for FG and SBP were associated with higher odds of LGA and SGA respectively (both P<0.01). While lower maternal FG and SBP are generally considered healthy in pregnancy, we found some evidence of association with higher odds of SGA (P=0.015) and LGA (P=0.14) respectively. We conclude that common genetic variants contribute to risk of SGA and LGA, but that additional factors become more important for risk of SGA in the smallest 3% of babies. Naturally low maternal glucose and blood pressure levels may additionally contribute to risk of SGA and LGA, respectively. Author SummaryBabies in the lowest or highest 10% of the population distribution of birth weight (BW) for a given gestational age are referred to as Small- or Large-for-Gestational-Age (SGA or LGA) respectively. These babies have higher risks of complications compared to babies with BW closer to the mean. SGA and LGA babies may have experienced growth restriction or overgrowth, respectively, but may alternatively just be at the tail ends of the normal growth distribution. The relative proportions of normal vs. sub-optimal growth within these groups is unclear. To examine the role of common genetic variation in SGA and LGA, we tested their associations with a fetal genetic score (GS) for BW in 12,125 European-ancestry individuals. We also tested associations with maternal GS (5,187 mothers) for offspring BW, fasting glucose and systolic blood pressure, each of which influences fetal growth via the in utero environment. We found all fetal and maternal GS were associated with SGA and LGA, supporting strong maternal and fetal genetic contributions to birth weight in both tails of the distribution. However, within the smallest 3% of babies, the maternal and fetal GS for BW were higher than expected, suggesting factors additional to common genetic variation are more important in determining birth weight in these very small babies.

genetics↗

bioRxiv2019

The effect of sex and underlying disease on the genetic association of QT interval and sudden cardiac death

BackgroundSudden cardiac death (SCD) accounts for ~300,000 deaths annually in the US. Men have a higher risk of SCD and are more likely to have underlying coronary artery disease (CAD) than women. In contrast, women are more likely to have arrhythmic events in the setting of inherited or acquired QT prolongation. Moreover, there is evidence of sex differences in the underlying genetics of QT interval duration. Using sex- and CAD-stratified analyses, we assess differences in genetic association between prolonged QT interval and SCD risk.\n\nMethodsWe examined 2,282 SCD subjects with autopsy-confirmed underlying disease from the Fingesture cohort and 3,561 Finnish controls. The SCD subjects were stratified by underlying disease (ischemic vs. non-ischemic) and by sex. We used logistic regression to test for association between the top QT interval associated SNP, rs12143842 (in the NOS1AP locus), and SCD risk. We also performed Mendelian randomization to test for causal association of QT interval in the various subgroups.\n\nResultsFemale SCD victims with underlying non-ischemic disease had the strongest association between rs12143842 and SCD risk (OR=1.37; 95% CI, 1.07-1.75) and the strongest causal association, established using Mendelian randomization, between prolonged QT interval and SCD (OR in SCD risk per SD increase in QT, 3.60; 95% CI, 1.22-10.49). Ischemic SCD victims, irrespective of sex, did not show an association between rs12143842 and SCD risk or a causal association for QT interval.\n\nConclusionsThis study provides evidence that the causal effect of QT prolongation on SCD risk differs by sex and underlying disease.

genetics↗

bioRxiv2019

methylSCOPA and META-methylSCOPA: software for the analysis and aggregation of epigenome-wide association studies of multiple correlated phenotypes

BackgroundMulti-phenotype genome-wide association studies (MP-GWAS) of correlated traits have greater power to detect genotype-phenotype associations than single-trait GWAS. However, no multi-phenotype analysis method exists for epigenome-wide association studies (EWAS).\n\nResultsWe extended the SCOPA approach developed by us to \"methylSCOPA\" software in C++ by reversely regressing DNA hyper/hypo-methylation information on a linear combination of phenotypes. We evaluated two models of association between DNA methylation and fasting glucose (FG) and insulin (FI) levels: Model 1, including FG, FI, and three measured potential confounders (body mass index [BMI], fasting serum triglyceride levels [TG], and waist/hip ratio [WHR]), and Model 2, including FG and FI corrected for the effects of BMI, TG, and WHR. Both models were additionally corrected for participant sex and smoking status (current/ever/never). We meta-analyzed the cohort-specific MP-EWAS results with our novel software META-methylSCOPA, mapped genomic locations to CGCh37/hg19, and adopted P<1x10-7 to denote epigenome-wide significance. We used the Illumina Infinium HumanMethylation450K BeadChip array data from the Northern Finland Birth Cohorts (NFBC) 1966/1986. We quality-controlled the data, regressed out the effects of measured potential confounders, and normalized the methylation signal intensity and FI data. The MP-EWAS included data for 643/457 individuals from NFBC1966 and NFBC1986, respectively (total N=1,100).\n\nIn Model 1, we detected epigenome-wide significant association in the MP-EWAS meta-analysis at cg13708645 (chr12:121,974,305; P=1.2x10-8) within KDM2B gene. Single-trait effects within KDM2B were on FI, BMI, and WHR. Model with effect on BMI and WHR showed the strongest association at this locus, while effect on FI in single-phenotype analysis was driven by the effect of adiposity. In Model 2, the strongest association was at cg05063096 (chr3:143,689,810; P=2.3x10-7) annotated to C3orf58 with strongest effect on FI in single-trait analysis and multi-phenotype effect on FI and WHI within Model 1.\n\nWe characterized the effects of established EWAS loci for diabetes and its risk factors and detected suggestive (p<0.01) associations at six markers including PHGDH, TXNIP, SLC7A11, CPT1A, MYO5C and ABCG1, through the dissection of the multi-phenotype effects in Model 1.\n\nConclusionsWe implemented MP-EWAS in methylSCOPA and demonstrated its enhanced power over single-trait EWAS for correlated phenotypes in large-scale data.

bioinformatics↗

bioRxiv2019

Genetic variation in apolipoprotein A-I concentrations and risk of coronary artery disease

RationaleApolipoprotein A-I (apoA-I) infusions represent a potential novel therapeutic approach for the prevention of coronary artery disease (CAD) with phase III cardiovascular outcome trials currently underway. Although circulating apoA-I levels inversely associate with risk of CAD, the evidence base of this representing a causal relationship is lacking.\n\nObjectiveTo assess the causal role of apoA-I in CAD using human genetics.\n\nMethods and ResultsWe identified a variant (rs12225230) in APOA1 locus that associated with circulating apoA-I concentrations at GWAS significance (P<5x10-8) in 20,370 Finnish participants and meta-analyzed our data with a previous genome-wide association study of apoA-I. We obtained genetic estimates of CAD from UK Biobank and CARDIoGRAMplusC4D (totaling 122,733 CAD cases) and conducted a two-sample Mendelian randomization analysis. We compared our genetic findings to observational associations of apoA-I with risk of CAD in 918 incident CAD cases among 11,535 individuals from population-based prospective cohorts. We also summarized the available evidence from randomized controlled trials (RCTs) of apoA-I infusion therapies reporting CAD events. ApoA-I was associated with a lower risk of CAD in observational analyses (HR 0.81; 95%CI: 0.75, 0.88; per 1-SD higher apoA-I), with the association showing a dose-response relationship. Rs12225230 associated with apoA-I concentrations (per-C allele beta 0.076 SD; SE: 0.013; P=1.5x10-9) but not with potential confounders. In Mendelian randomization analyses, apoA-I was not related to risk of CAD (OR 1.13; 95%CI: 0.98, 1.30 per 1-SD higher apoA-I), which was different to the observational association (P-het<0.001). RCTs of apoA-I infusions did not show an effect on the risk of CAD.\n\nConclusionsGenetic evidence fails to support a cardioprotective role for apoA-I. This casts doubt on the likely benefit of apoA-I infusion therapy in the ongoing phase III cardiovascular outcome trial.

epidemiology↗

bioRxiv2019

Circulating metabolites and the risk of type 2 diabetes: a prospective study of 11,896 young adults from four Finnish cohorts

ObjectiveAdvances in metabolomics now allow high-throughput biomarker profiling of large population studies. We aimed to identify circulating metabolic biomarkers predictive of type 2 diabetes in young adults.\n\nMethodsNuclear magnetic resonance metabolomics was used to quantify 229 metabolic measures in 11,896 individuals from four Finnish cohorts (mean age 33 years, range 24-45). Associations between baseline metabolites and risk of type 2 diabetes onset during 8-15 years of follow-up (392 incident cases) were assessed by logistic regression adjusted for sex, age, body mass index, and fasting glucose.\n\nResultsOut of 229 metabolic measures, 113 were associated with incident diabetes in meta-analysis of the four cohorts (P<0.0009; odds ratios per 1-SD: 0.59-1.50). Among the strongest predictors of diabetes risk were branched-chained and aromatic amino acids (odds ratios 1.31-1.33), triglyceride fractions within the largest very-low-density lipoprotein particles (VLDL; odds ratios 1.33-1.50)), as well as linoleic omega-6 fatty acids (odds ratio 0.75) and free cholesterol in large high-density lipoprotein particles (HDL; odds ratio 0.59). A biomarker score comprised of phenylalanine, free cholesterol in large HDL, and the ratio of cholesteryl esters to total lipids in large VLDL was predictive of the risk for future diabetes in an independent validation cohort (odds ratio 10.1 [95% confidence intervals 4.2-24.1] comparing individuals in upper vs lower fifth of biomarker score). Adjustment for routine lipids and insulin attenuated the odds ratio to 5.8 [2.2-15.1].\n\nConclusionsMetabolic aberrations across multiple molecular pathways are predictive of the long-term risk of type 2 diabetes in young adults. Comprehensive metabolic profiling may potentially help targeting preventive interventions for young asymptomatic individuals at increased risk for type 2 diabetes.

epidemiology↗

bioRxiv2018

Genome-wide association study identifies seven novel loci associating with circulating cytokines and cell adhesion molecules in Finns

BackgroundInflammatory processes contribute to the pathophysiology of multiple chronic conditions. Genetic factors play a crucial role in modulating the inflammatory load, but the exact mechanisms are incompletely understood.\n\nMethodsTo add understanding to the molecular mechanisms in inflammation, we performed a genome-wide association study (GWAS) on 16 circulating cytokines and cell adhesion molecules (inflammatory phenotypes) in Northern Finland Birth Cohort 1966 (NFBC1966, N=5,284). A subsequent meta-analysis was completed for 10 phenotypes available in a GWAS of three other Finnish population cohorts adding up to 13,577 individuals in the study. Complementary association tests were performed to study the effect of the ABO blood types on soluble adhesion molecule levels.\n\nResultsWe identified seven novel and confirmed six previously reported loci associating with at least one of the studied inflammatory phenotypes (p<3.1x10-9). We observed three loci associating with the concentration of soluble vascular cell adhesion molecule-1 (sVCAM-1), one of which is the ABO locus that has been previously associated with soluble E-selectin (sE-selectin) and intercellular adhesion molecule-1 (sICAM-1) levels. Results from the complementary analyses suggest that the blood type B associates primarily with the concentration of sVCAM-1 while the A1 subtype shows a robust effect on sE-selectin and sICAM-1 levels. Furthermore, the genotypes in the ABO locus associating with higher soluble adhesion molecule levels tend to associate with lower low-density lipoprotein cholesterol level and lower cardiovascular disease risk.\n\nConclusionThe present results extend the knowledge about genetic factors contributing to the inflammatory load. Our findings suggest that two distinct mechanisms contribute to the soluble adhesion molecule levels at the ABO locus. The negative correlation between the genetic effects on soluble adhesion molecule levels and cardiovascular traits in this locus further suggests that increased soluble adhesion molecule levels per se may not be a risk factor for cardiovascular disease.

genetics↗

bioRxiv2018

Lower hemoglobin levels associate with lower body mass index and healthier metabolic profile

Hemoglobin (Hb) is the main carrier of oxygen. In general, high-end Hb levels within the normal range are considered beneficial for health1. However, activation of the hypoxia response has been shown to protect mice against metabolic dysfunction2-4. We used Hb levels as an indicator for oxygenation status and studied its association with >170 anthropometric and metabolic parameters in two Finnish birth cohorts both in cross-sectional and longitudinal design (max n = 7,175). Here we show a positive linear association between Hb levels and body mass index (BMI). Subjects with the lower Hb levels had better glucose tolerance, lower cholesterol and blood pressure levels, less adverse metabolite profiles and lower inflammatory load. Notably, these associations were not only mediated by the lower BMI, and the effect size of many of them increased with age. Polygenic risk score (PRS) analyses indicated shared genetic determinants between Hb levels and BMI, insulin, triglyceride and HDL cholesterol levels. Mendelian randomization (MR) analyses could not demonstrate causal relationships between Hb and metabolic parameters. However, manipulation of Hb levels by venesection in mice showed evidence for causal associations with body weight and metabolic parameters. Our findings suggest that lower-end normal Hb levels may be favorable for systemic metabolism involving mild chronic activation of the hypoxia response. Therefore modulation of Hb levels could be a novel strategy towards maintenance of metabolic health.

epidemiology↗

bioRxiv2018

Sparse variable and covariance selection for high-dimensional seemingly unrelated Bayesian regression

MotivationOur work is motivated by the search for metabolite Quantitative Trait Loci (QTL) in a cohort of more than 5,000 people. There are 158 metabolites measured by NMR spectroscopy measured in the 31-year follow-up of the Northern Finland Birth Cohort 1966 (NFBC66). These metabolites, as with many multivariate phenotypes produced by high-throughput biomarker technology, exhibit strong correlation structures. Existing approaches for combining such data with genetic variants for multivariate QTL analysis generally ignore phenotypic correlations or make restrictive assumptions about the associations between phenotypes and genetic loci. ResultsWe present a computationally efficient Bayesian Seemingly Unrelated Regressions (SUR) model for high-dimensional data, with cell-sparse variable selection and sparse graphical structure for covariance selection. Cell-sparsity allows different phenotype responses to be associated with different genetic predictors and the graphical structure is used to represent the conditional dependencies between phenotype variables. To achieve feasible computation of the large model space, we exploit a factorisation of the covariance matrix. Applying the model to the NFBC66 data with 9,000 directly-genotyped Single Nucleotide Polymorphisms, we are able to simultaneously estimate genotype-phenotype associations and the residual dependence structure amongst the metabolites at the same time. Availability and implementationThe R package BayesSUR with full documentation is available at https://cran.r-project.org/web/packages/BayesSUR/ Contactalex.lewin@lshtm.ac.uk

bioinformatics↗

bioRxiv2018

Exome sequencing identifies high impact trait-associated alleles enriched in Finns

As yet undiscovered rare variants are hypothesized to substantially influence an individuals risk for common diseases and traits, but sequencing studies aiming to identify such variants have generally been underpowered. In isolated populations that have expanded rapidly after a population bottleneck, deleterious alleles that passed through the bottleneck may be maintained at much higher frequencies than in other populations. In an exome sequencing study of nearly 20,000 cohort participants from northern and eastern Finnish populations that exemplify this phenomenon, most novel trait-associated deleterious variants displayed frequencies 10-173 times higher than in other European populations. These enriched alleles underlie 30 novel associations with 20 disease-related quantitative traits and demonstrate a geographical clustering equivalent to that of Mendelian disease mutations characteristic of the Finnish population. Sequencing studies in populations without this unique history would require hundreds of thousands to millions of participants for comparable power.

genomics↗

bioRxiv2018

DNA methylation links prenatal smoking exposure to later life health outcomes in offspring.

Maternal smoking during pregnancy is associated with adverse offspring health outcomes across their life course. We hypothesize that DNA methylation is a potential mediator of this relationship. To test this, we examined the association of prenatal maternal smoking with DNA methylation in 2,821 individuals (age 16 to 48 years) from five prospective birth cohort studies and perform Mendelian randomization and mediation analyses to assess, whether methylation markers have causal effects on disease outcomes in the offspring. We identify 69 differentially methylated CpGs in 36 genomic regions (P < 1x10-7), and show that DNA methylation may represent a biological mechanism through which maternal smoking is associated with increased risk of psychiatric morbidity in the exposed offspring.

epidemiology↗

bioRxiv2018

Determinants of accelerated metabolomic and epigenetic ageing in a UK cohort

Markers of biological ageing have potential utility in primary care and public health. We developed an elastic net regression model of age based on untargeted metabolic profiling across multiple platforms, including nuclear magnetic resonance spectroscopy and liquid chromatography-mass spectrometry in urine and serum (almost 100,000 features assayed), within a large sample (N=2,239) from the UK occupational Airwave cohort. We investigated the determinants of accelerated ageing, including genetic, lifestyle and psychological risk factors for premature mortality. The metabolomic age model was well correlated with chronological age (r=0.85 in independent test set). Increased metabolomic age acceleration (mAA) was associated (p<0.0025) with overweight/obesity and depression and nominally associated (p<0.05) with high alcohol use and low income. DNA methylation age acceleration (N=1,102) was nominally associated (p<0.05) with high alcohol use, anxiety and post-traumatic stress disorder, but not correlated with mAA. Biological age acceleration may present an important mechanism linking psycho-social stress to age-related disease.

epidemiology↗

bioRxiv2018

Large-scale genome-wide association meta-analysis of endometriosis reveals 13 novel loci and genetically-associated comorbidity with other pain conditions

Endometriosis is a common complex inflammatory condition characterised by the presence of endometrium-like tissue outside the uterus, mainly in the pelvic area. It is associated with chronic pelvic pain and infertility, and its pathogenesis remains poorly understood. The disease is typically classified according to the revised American Fertility Society (rAFS) 4-stage surgical assessment system, although stage does not correlate well with symptomatology or prognosis. Previously identified genetic variants mainly are associated with stage III/IV disease, highlighting the need for further phenotype-stratified analysis that requires larger datasets. We conducted a meta-analysis of 15 genome-wide association studies (GWAS) and a replication analysis, including 58,115 cases and 733,480 controls in total, and sub-phenotype analyses of stage I/II, stage III/IV and infertility-associated endometriosis cases. This revealed 27 genetic loci associated with endometriosis at the genome-wide p-value threshold (P<5x10-8), 13 of which are novel and an additional 8 novel genes identified from gene-based association analyses. Of the 27 loci, 21 (78%) had greater effect sizes in stage III/IV disease compared to stage I/II, 1 (4%) had greater effect size in stage I/II compared to stage III/IV and 17 (63%) had greater effect sizes when restricted to infertility-associated endometriosis cases compared to overall endometriosis. These results suggest that specific variants may confer risk for different sub-types of endometriosis through distinct pathways. Analyses of genetic variants underlying different pain symptoms reported in the UK Biobank showed that 7/9 had positive significant (p<1.28x103) positive genetic correlations with endometriosis, suggesting a genetic basis for sensitivity to pain in general. Additional conditions with significant positive genetic correlations with endometriosis included uterine fibroids, excessive and irregular menstrual bleeding, osteoarthritis, diabetes as well as menstrual cycle length and age at menarche. These results provide a basis for fine-mapping of the causal variants at these 27 loci, and for functional follow-up to understand their contribution to endometriosis and its potential subtypes.

genomics↗

bioRxiv2018

Machine Learning in Multi-Omics Data to Assess Longitudinal Predictors of Glycaemic Trait Levels

Type 2 diabetes (T2D) is a global health burden that will benefit from personalised risk prediction and targeted prevention programmes. Omics data have enabled more detailed risk prediction; however, most studies have focussed on directly on the ability of DNA variants predicting T2D onset with less attention given to epigenetic regulation and glycaemic trait variability. By applying machine learning to the longitudinal Northern Finland Birth Cohort 1966 (NFBC 1966) at 31 (T1) and 46 (T2) years old, we predicted fasting glucose (FG) and insulin (FI), glycated haemoglobin (HbA1c) and 2-hour glucose and insulin from oral glucose tolerance test (2hGlu, 2hIns) at T2 in 513 individuals from 1,001 variables at T1 and T2, including anthropometric, metabolic, metabolomic and epigenetic variables. We further tested whether the information obtained by the machine learning models in NFBC could be used to predict glycaemic traits in the independent French study with 48 matching predictors (DESIR, N=769, age range 30-65 years at recruitment, interval between data collections: 9 years). In this study, FG and FI were best predicted, with average R2 values of 0.38 and 0.53. Sex, branched-chain and aromatic amino acids, HDL-cholesterol, glycerol, ketone bodies, blood pressure at T2 and measurements of adiposity at T1, as well as multiple methylation marks at both time points were amongst the top predictors. In the validation analysis, we reached R2 values of 0.41/0.55 for FG/FI when trained and tested in NFBC1966 and 0.17/0.30 when trained in NFBC1966 and tested in DESIR. We identified clinically relevant sets of predictors from a large multi-omics dataset and highlighted the potential of methylation markers and longitudinal changes in prediction.

genomics↗

bioRxiv2018

Genome-wide association analysis identifies 27 novel loci associated with uterine leiomyomata revealing common genetic origins with endometriosis

Uterine leiomyomata (UL), also known as uterine fibroids, are the most common neoplasms of the reproductive tract and the primary cause for hysterectomy, leading to considerable impact on womens lives as well as high economic burden1,2. Genetic epidemiologic studies indicate that heritable risk factors contribute to UL pathogenesis3. Previous genome-wide association studies (GWAS) identified five loci associated with UL at genome-wide significance (P < 5 x 10-8)4-6. We conducted GWAS meta-analysis in 20,406 cases and 223,918 female controls of white European ancestry, identifying 24 genome-wide significant independent loci; 17 replicated in an unrelated cohort of 15,068 additional cases and 43,587 female controls. Aggregation of discovery and replication studies (35,474 cases and 267,505 female controls) revealed six additional significant loci. Interestingly, four of the 17 loci identified and replicated in these analyses have also been associated with risk for endometriosis - another common gynecologic disorder. These findings increase our understanding of the biological mechanisms underlying UL development, and suggest overlapping genetic origins with endometriosis.

genomics↗

bioRxiv2018

Large-Scale Genome-Wide Meta Analysis of Polycystic Ovary Syndrome Suggests Shared Genetic Architecture for Different Diagnosis Criteria.

Polycystic ovary syndrome (PCOS) is a disorder characterized by hyperandrogenism, ovulatory dysfunction and polycystic ovarian morphology. Affected women frequently have metabolic disturbances including insulin resistance and dysregulation of glucose homeostasis. PCOS is diagnosed with two different sets of diagnostic criteria, resulting in a phenotypic spectrum of PCOS cases. The genetic similarities between cases diagnosed with different criteria have been largely unknown. Previous studies in Chinese and European subjects have identified 16 loci associated with risk of PCOS. We report a meta-analysis from 10,074 PCOS cases and 103,164 controls of European ancestry and characterisation of PCOS related traits. We identified 3 novel loci (near PLGRKT, ZBTB16 and MAPRE1), and provide replication of 11 previously reported loci. Identified variants were associated with hyperandrogenism, gonadotropin regulation and testosterone levels in affected women. Genetic correlations with obesity, fasting insulin, type 2 diabetes, lipid levels and coronary artery disease indicate shared genetic architecture between metabolic traits and PCOS. Mendelian randomization analyses suggested variants associated with body mass index, fasting insulin, menopause timing, depression and male-pattern balding play a causal role in PCOS. Only one locus differed in its association by diagnostic criteria, otherwise the genetic architecture was similar between PCOS diagnosed by self-report and PCOS diagnosed by NIH or Rotterdam criteria across common variants at 13 loci.

genetics↗