[Show abstract][Hide abstract] ABSTRACT: Background
Statins increase the risk of new-onset type 2 diabetes mellitus. We aimed to assess whether this increase in risk is a consequence of inhibition of 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), the intended drug target.
We used single nucleotide polymorphisms in the HMGCR gene, rs17238484 (for the main analysis) and rs12916 (for a subsidiary analysis) as proxies for HMGCR inhibition by statins. We examined associations of these variants with plasma lipid, glucose, and insulin concentrations; bodyweight; waist circumference; and prevalent and incident type 2 diabetes. Study-specific effect estimates per copy of each LDL-lowering allele were pooled by meta-analysis. These findings were compared with a meta-analysis of new-onset type 2 diabetes and bodyweight change data from randomised trials of statin drugs. The effects of statins in each randomised trial were assessed using meta-analysis.
Data were available for up to 223 463 individuals from 43 genetic studies. Each additional rs17238484-G allele was associated with a mean 0·06 mmol/L (95% CI 0·05–0·07) lower LDL cholesterol and higher body weight (0·30 kg, 0·18–0·43), waist circumference (0·32 cm, 0·16–0·47), plasma insulin concentration (1·62%, 0·53–2·72), and plasma glucose concentration (0·23%, 0·02–0·44). The rs12916 SNP had similar effects on LDL cholesterol, bodyweight, and waist circumference. The rs17238484-G allele seemed to be associated with higher risk of type 2 diabetes (odds ratio [OR] per allele 1·02, 95% CI 1·00–1·05); the rs12916-T allele association was consistent (1·06, 1·03–1·09). In 129 170 individuals in randomised trials, statins lowered LDL cholesterol by 0·92 mmol/L (95% CI 0·18–1·67) at 1-year of follow-up, increased bodyweight by 0·24 kg (95% CI 0·10–0·38 in all trials; 0·33 kg, 95% CI 0·24–0·42 in placebo or standard care controlled trials and −0·15 kg, 95% CI −0·39 to 0·08 in intensive-dose vs moderate-dose trials) at a mean of 4·2 years (range 1·9–6·7) of follow-up, and increased the odds of new-onset type 2 diabetes (OR 1·12, 95% CI 1·06–1·18 in all trials; 1·11, 95% CI 1·03–1·20 in placebo or standard care controlled trials and 1·12, 95% CI 1·04–1·22 in intensive-dose vs moderate dose trials).
The increased risk of type 2 diabetes noted with statins is at least partially explained by HMGCR inhibition.
The funding sources are cited at the end of the paper.
[Show abstract][Hide abstract] ABSTRACT: Previous population-based studies of plasma ferritin concentration have not revealed a relationship with total mortality. We tested the possible association of increased ferritin concentrations with increased risk of total and cause-specific mortality in the general population.
[Show abstract][Hide abstract] ABSTRACT: Homozygous familial hypercholesterolaemia (HoFH) is a rare life-threatening condition characterized by markedly elevated circulating levels of low-density lipoprotein cholesterol (LDL-C) and accelerated, premature atherosclerotic cardiovascular disease (ACVD). Given recent insights into the heterogeneity of genetic defects and clinical phenotype of HoFH, and the availability of new therapeutic options, this Consensus Panel on Familial Hypercholesterolaemia of the European Atherosclerosis Society (EAS) critically reviewed available data with the aim of providing clinical guidance for the recognition and management of HoFH.
[Show abstract][Hide abstract] ABSTRACT: Background High plasma levels of nonfasting triglycerides are associated with an increased risk of ischemic cardiovascular disease. Whether lifelong low levels of nonfasting triglycerides owing to mutations in the gene encoding apolipoprotein C3 (APOC3) are associated with a reduced risk of ischemic cardiovascular disease in the general population is unknown. Methods Using data from 75,725 participants in two general-population studies, we first tested whether low levels of nonfasting triglycerides were associated with reduced risks of ischemic vascular disease and ischemic heart disease. Second, we tested whether loss-of-function mutations in APOC3, which were associated with reduced levels of nonfasting triglycerides, were also associated with reduced risks of ischemic vascular disease and ischemic heart disease. During follow-up, ischemic vascular disease developed in 10,797 participants, and ischemic heart disease developed in 7557 of these 10,797 participants. Results Participants with nonfasting triglyceride levels of less than 1.00 mmol per liter (90 mg per deciliter) had a significantly lower incidence of cardiovascular disease than those with levels of 4.00 mmol per liter (350 mg per deciliter) or more (hazard ratio for ischemic vascular disease, 0.43; 95% confidence interval [CI], 0.35 to 0.54; hazard ratio for ischemic heart disease, 0.40; 95% CI, 0.31 to 0.52). Heterozygosity for loss-of-function mutations in APOC3, as compared with no APOC3 mutations, was associated with a mean reduction in nonfasting triglyceride levels of 44% (P<0.001). The cumulative incidences of ischemic vascular disease and ischemic heart disease were reduced in heterozygotes as compared with noncarriers of APOC3 mutations (P=0.009 and P=0.05, respectively), with corresponding risk reductions of 41% (hazard ratio, 0.59; 95% CI, 0.41 to 0.86; P=0.007) and 36% (hazard ratio, 0.64; 95% CI, 0.41 to 0.99; P=0.04). Conclusions Loss-of-function mutations in APOC3 were associated with low levels of triglycerides and a reduced risk of ischemic cardiovascular disease. (Funded by the European Union and others.).
New England Journal of Medicine 06/2014; · 54.42 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Markers of the chromosome 9p21 region are regarded as the strongest and most reliably significant genome-wide association study (GWAS) signals for Coronary heart disease (CHD) risk; this was recently confirmed by the CARDIoGRAMplusC4D Consortium meta-analysis. However, while these associations are significant at the population level, they may not be clinically relevant predictors of risk for all individuals. We describe here the results of a study designed to address the question: What is the contribution of context defined by traditional risk factors in determining the utility of DNA sequence variations marking the 9p21 region for explaining variation in CHD risk? We analyzed a sample of 7,589 (3,869 females and 3,720 males) European American participants of the Atherosclerosis Risk in Communities study. We confirmed CHD-SNP genotype associations for two 9p21 region marker SNPs previously identified by the CARDIoGRAMplusC4D Consortium study, of which ARIC was a part. We then tested each marker SNP genotype effect on prediction of CHD within sub-groups of the ARIC sample defined by traditional CHD risk factors by applying a novel multi-model strategy, PRIM. We observed that the effects of SNP genotypes in the 9p21 region were strongest in a sub-group of hypertensives. We subsequently validated the effect of the region in an independent sample from the Copenhagen City Heart Study. Our study suggests that marker SNPs identified as predictors of CHD risk in large population based GWAS may have their greatest utility in explaining risk of disease in particular sub-groups characterized by biological and environmental effects measured by the traditional CHD risk factors.
[Show abstract][Hide abstract] ABSTRACT: Long QT syndrome (LQTS) is a cardiac ion channelopathy which presents clinically with palpitations, syncope or sudden death. More than 700 LQTS-causing mutations have been identified in 13 genes, all of which encode proteins involved in the execution of the cardiac action potential. The most frequently affected genes, covering > 90% of cases, are KCNQ1, KCNH2 and SCN5A.
We describe 64 different mutations in 70 unrelated Danish families using a routine five-gene screen, comprising KCNQ1, KCNH2 and SCN5A as well as KCNE1 and KCNE2.
Twenty-two mutations were found in KCNQ1, 28 in KCNH2, 9 in SCN5A, 3 in KCNE1 and 2 in KCNE2. Twenty-six of these have only been described in the Danish population and 18 are novel. One double heterozygote (1.4% of families) was found. A founder mutation, p.F29L in KCNH2, was identified in 5 "unrelated" families. Disease association, in 31.2% of cases, was based on the type of mutation identified (nonsense, insertion/deletion, frameshift or splice-site). Functional data was available for 22.7% of the missense mutations. None of the mutations were found in 364 Danish alleles and only three, all functionally characterised, were recorded in the Exome Variation Server, albeit at a frequency of < 1:1000.
The genetic etiology of LQTS in Denmark is similar to that found in other populations. A large founder family with p.F29L in KCNH2 was identified. In 48.4% of the mutations disease causation was based on mutation type or functional analysis.
BMC Medical Genetics 03/2014; 15(1):31. · 2.54 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Nonalcoholic fatty liver disease (NAFLD) is the most common form of liver disease. To elucidate the molecular basis of NAFLD, we performed an exome-wide association study of liver fat content. Three variants were associated with higher liver fat levels at the exome-wide significance level of 3.6 × 10(-7): two in PNPLA3, an established locus for NAFLD, and one (encoding p.Glu167Lys) in TM6SF2, a gene of unknown function. The TM6SF2 variant encoding p.Glu167Lys was also associated with higher circulating levels of alanine transaminase, a marker of liver injury, and with lower levels of low-density lipoprotein-cholesterol (LDL-C), triglycerides and alkaline phosphatase in 3 independent populations (n > 80,000). When recombinant protein was expressed in cultured hepatocytes, 50% less Glu167Lys TM6SF2 protein was produced relative to wild-type TM6SF2. Adeno-associated virus-mediated short hairpin RNA knockdown of Tm6sf2 in mice increased liver triglyceride content by threefold and decreased very-low-density lipoprotein (VLDL) secretion by 50%. Taken together, these data indicate that TM6SF2 activity is required for normal VLDL secretion and that impaired TM6SF2 function causally contributes to NAFLD.
[Show abstract][Hide abstract] ABSTRACT: Increased nonfasting plasma triglycerides marking increased amounts of cholesterol in remnant lipoproteins are important risk factors for cardiovascular disease, but whether lifelong reduced concentrations of triglycerides on a genetic basis ultimately lead to reduced all-cause mortality is unknown. We tested this hypothesis.METHODS: Using individuals from the Copenhagen City Heart Study in a Mendelian randomization design, we first tested whether low concentrations of nonfasting triglycerides were associated with reduced all-cause mortality in observational analyses (n = 13 957); second, whether genetic variants in the triglyceride-degrading enzyme lipoprotein lipase, resulting in reduced nonfasting triglycerides and remnant cholesterol, were causally associated with reduced all-cause mortality (n = 10 208).RESULTS: During a median 24 and 17 years of 100% complete follow-up, 9991 and 4005 individuals died in observational and genetic analyses, respectively. In observational analyses compared to individuals with nonfasting plasma triglycerides of 266-442 mg/dL (3.00-4.99 mmol/L), multivariably adjusted hazard ratios for all-cause mortality were 0.89 (95% CI 0.78-1.02) for 177-265 mg/dL (2.00-2.99 mmol/L), 0.74 (0.65-0.84) for 89-176 mg/dL (1.00-1.99 mmol/L), and 0.59 (0.51-0.68) for individuals with nonfasting triglycerides <89 mg/dL (<1.00 mmol/L). The odds ratio for a genetically derived 89-mg/dL (1-mmol/L) lower concentration in nonfasting triglycerides was 0.50 (0.30-0.82), with a corresponding observational hazard ratio of 0.87 (0.85-0.89). Also, the odds ratio for a genetically derived 50% lower concentration in nonfasting triglycerides was 0.43 (0.23-0.80), with a corresponding observational hazard ratio of 0.73 (0.70-0.77).CONCLUSIONS: Genetically reduced concentrations of nonfasting plasma triglycerides are associated with reduced all-cause mortality, likely through reduced amounts of cholesterol in remnant lipoproteins.
[Show abstract][Hide abstract] ABSTRACT: Objective
To test the hypothesis that genetic variation in ABCG5/8, the transporter responsible for intestinal and hepatobiliary cholesterol efflux, might simultaneously influence plasma and biliary cholesterol levels, and hence risk of myocardial infarction(MI) and gallstone disease in opposite directions.
High plasma levels of low-density lipoprotein(LDL) cholesterol are a causal risk factor for MI, whereas high levels of biliary cholesterol promote gallstone formation.
A total of 60,239 subjects from Copenhagen were included, including 5,647 with MI and 3,174 with symptomatic gallstone disease. Subjects were genotyped for six common, nonsynonymous and functional variants in ABCG5/8, and a combined weighted genotype score was calculated.
Combined, weighted genotype scores were associated with stepwise decreases in LDL cholesterol of up to 5.9%(0.20 mmol/L) for individuals with a score ≥8.0 (prevalence=11%) versus <2.0 (prevalence=9%) (p for trend across five groups=2×10E-35). The cumulative incidences of myocardial infarction and gallstone disease as a function of age and increasing genotype score were, respectively, decreased and increased(log-rank p-trend: 6×10E-4 and 9×10E-45). The multifactorially adjusted odds ratios were 0.83(95% confidence interval, 0.73 to 0.94) for myocardial infarction and 2.85(2.39 to 3.39) for symptomatic gallstone disease for individuals with a genotype score ≥8.0 versus <2.0.
Genetic variation in ABCG5/8 which associates with decreased levels of plasma LDL cholesterol protects against myocardial infarction, but increases the risk of symptomatic gallstone disease. These results suggest that myocardial infarction and gallstones, two seemingly unrelated diseases, are intrinsically linked via the function of the ABCG5/8 cholesterol transporter.
Journal of the American College of Cardiology 01/2014; · 14.09 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Plasma triglyceride concentration is a biomarker for circulating triglyceride-rich lipoproteins and their metabolic remnants. Common mild-to-moderate hypertriglyceridaemia is typically multigenic, and results from the cumulative burden of common and rare variants in more than 30 genes, as quantified by genetic risk scores. Rare autosomal recessive monogenic hypertriglyceridaemia can result from large-effect mutations in six different genes. Hypertriglyceridaemia is exacerbated by non-genetic factors. On the basis of recent genetic data, we redefine the disorder into two states: severe (triglyceride concentration >10 mmol/L), which is more likely to have a monogenic cause; and mild-to-moderate (triglyceride concentration 2-10 mmol/L). Because of clustering of susceptibility alleles and secondary factors in families, biochemical screening and counselling for family members is essential, but routine genetic testing is not warranted. Treatment includes management of lifestyle and secondary factors, and pharmacotherapy. In severe hypertriglyceridaemia, intervention is indicated because of pancreatitis risk; in mild-to-moderate hypertriglyceridaemia, intervention can be indicated to prevent cardiovascular disease, dependent on triglyceride concentration, concomitant lipoprotein disturbances, and overall cardiovascular risk.
[Show abstract][Hide abstract] ABSTRACT: Cardiovascular disease is one of the most common age-related diseases, and also one of the most common causes of death in the general population. We tested the hypothesis that visible age-related signs associate with risk of ischemic heart disease (IHD), myocardial infarction (MI), and death in the general population, independent of chronological age.
10,885 individuals aged 20-93 years free of IHD were followed from 1976-8 until June 2011 with 100% complete follow-up. During these 35 years of follow-up, 3,401 participants developed IHD and 1,708 developed MI. Presence of frontoparietal baldness, crown top baldness, earlobe crease, and xanthelasmata was associated with increased risk of IHD and/or MI after multifactorial adjustment for chronological age and well-known cardiovascular risk factors. The risk of IHD and MI increased stepwise with increasing number of age-related signs with multifactorially adjusted hazard ratios up to 1.40 (95% confidence interval 1.20 to 1.62) for IHD and 1.57 (1.28 to 1.93) for MI, in individuals with 3-4 versus no age-related signs at baseline (P for trend <0.001). In all age-groups in both women and men, absolute 10-year risk of IHD and MI increased with increasing number of visible age-related signs.
Male pattern baldness, earlobe crease, and xanthelasmata - alone or in combination - associate with increased risk of ischemic heart disease and myocardial infarction independent of chronological age and other well-known cardiovascular risk factors. This is the first prospective study to show that looking old for your age is a marker of poor cardiovascular health.
[Show abstract][Hide abstract] ABSTRACT: Use of oral contraceptives with estrogen and hormone replacement therapy with estrogen or testosterone are associated with increased risk of venous thromboembolism (VTE). However, whether endogenous estradiol and testosterone concentrations also associate with risk of VTE is unknown.
We tested the hypothesis that elevated endogenous total estradiol and total testosterone concentrations are associated with increased risk of VTE in the general population.
We studied 4,658 women, not receiving exogenous estrogen, and 4,673 men from the 1981-1983 examination of the Copenhagen City Heart Study, who had estradiol and testosterone concentrations measured. Of these, 636 developed VTE (deep venous thrombosis (DVT) and/or pulmonary embolism (PE)) during a follow-up of 21 years (range: 0.02-32years). Associations between endogenous estradiol and testosterone concentrations and risk of VTE were estimated by Cox proportional hazards regression with time dependent covariates and corrected for regression dilution bias.
Multifactorially adjusted hazard ratios of VTE for individuals with estradiol >75(th) versus ≤25(th) percentile were 0.84(95%CI: 0.25-2.85), 1.05(0.53-2.08), and 1.05(0.03-35.13) for pre-, and post-menopausal women and men, respectively. For testosterone, corresponding risk estimates were 0.64(0.03-12.32), 1.11(0.66-1.86), and 1.30(0.62-2.73). In addition, no associations were observed for extreme hormone percentiles (>95(th) versus ≤75(th) ) and risk of DVT and PE or recurrent VTE.
This prospective study suggests that high endogenous concentrations of estradiol and testosterone in women and men in the general population are not associated with increased risk of VTE, DVT or PE. This article is protected by copyright. All rights reserved.
Journal of Thrombosis and Haemostasis 12/2013; · 6.08 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Objective
Mortality is increased in patients with hereditary hemochromatosis, in individuals from the general population with increased transferrin saturation(TS), and also in patients with diabetes type 1 and increased TS from a highly specialised diabetes clinic. Thus, we have recommended targeted screening for TS in specialised diabetes clinics. Whether mortality is also increased in individuals ascertained from the general population with diabetes and increased TS is unknown.Research design and methodsIn two Danish population studies(N=84,865), we examined mortality according to baseline TS and hemochromatosis genotype(HFE) C282Y/C282Y in individuals with diabetes(type 1,N=118;type 2,N=3228;total,N=3346).ResultsThe cumulative survival was reduced in individuals with diabetes with TS≥50% vs.<50%(log-rank;P<0.0001),with median survival age of 66 and 79 years, respectively. The hazard ratio(HR) for TS≥50% vs.<50% was 2.0(95%CI:1.3-2.8;P=0.0004) for total mortality overall(and similar for men and women separately); 2.6(1.3-5.4;P=0.008) for neoplasms and 3.4(2.0-6.0;P=0.00002) for endocrinological causes. A stepwise increased risk of total mortality was observed for stepwise increasing levels of TS(log-rank:P=0.0001), with a HR for TS≥70% vs.TS<20% of 4.8(2.0-12;P =0.0006). The HR for total mortality in individuals with diabetes for C282Y/C282Y vs. wildtype/wildtype was 3.3(1.04-10;P=0.04), and for (C282Y/C282Y&TS≥50%) vs. (wildtype/wildtype&TS<50%) was 6.0(1.5-24;P=0.01). Six percent of these premature deaths can possibly be avoided by early screening for TS or HFE genotype.Conclusions
Individuals with diabetes, ascertained in the general population, with increased TS or HFE genotype have a 2-6-fold increased risk of premature death.
[Show abstract][Hide abstract] ABSTRACT: To determine whether elevated lipoprotein(a) levels and corresponding LPA risk genotypes (rs10455872, rs3798220, kringle IV type 2 repeat polymorphism) prospectively associate with increased risk of aortic valve stenosis (AVS).
The etiologic basis of AVS is unclear. Recent data implicate an LPA genetic variant (rs10455872), associated with lipoprotein(a) levels, in calcific AVS.
We combined data from two prospective general population studies, the Copenhagen City Heart Study (1991-2011;N=10803) and the Copenhagen General Population Study (2003-2011;N=66877), following 77680 Danish participants for up to 20 years, during which time 454 were diagnosed with AVS. We conducted observational and genetic instrumental variable analyses in a Mendelian randomization study design.
Elevated lipoprotein(a) levels associated with multivariable adjusted hazard ratios for AVS of 1.2(95% confidence interval:0.8-1.7) for 22(nd) to 66(th) percentile levels (4-15 mg/dL), 1.6(1.1-2.4) for 67(th) to 89(th) percentile levels (16-55 mg/dL), 2.0(1.2-3.4) for 90(th) to 95(th) percentile levels (56-74 mg/dL), and 2.9(1.8-4.9) for levels >95(th) percentile (>74 mg/dL), versus levels <22(nd) percentile (<4mg/dL) (trend, p<0.001). Lipoprotein(a) levels were elevated among carriers of rs10455872 and rs3798220 minor alleles, and of low number of KIV-2 repeats (trend, all p<0.001). Combining all genotypes, instrumental variable analysis yielded a genetic relative risk for AVS of 1.6(95% CI, 1.2-2.1) for a 10-fold lipoprotein(a) increase, comparable to the observational hazard ratio of 1.4(1.2-1.7) for a 10-fold increase in lipoprotein(a) plasma levels.
Elevated lipoprotein(a) levels and corresponding genotypes associated with increased risk of AVS in the general population, with levels >74 mg/dL predicting a 3-fold increased risk.
Journal of the American College of Cardiology 10/2013; · 14.09 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Elevated nonfasting remnant cholesterol and low-density lipoprotein(LDL) cholesterol are causally associated with ischemic heart disease(IHD), but whether elevated nonfasting remnant cholesterol and LDL cholesterol both cause low-grade inflammation is presently unknown.
We studied 60,608 individuals from the Copenhagen General Population Study, the Copenhagen City Heart Study, and the Copenhagen Ischemic Heart Disease study, of which 10,668 had IHD diagnosed between 1977 and 2011. We genotyped for variants affecting levels of 1) nonfasting remnant cholesterol, 2) LDL cholesterol, 3) C-reactive protein by CRP alleles, and 4) C-reactive protein by interleukin 6 receptor alleles. Using a multi-directional Mendelian randomization design, we investigated possible causal associations between the lipoproteins and C-reactive protein, and between the lipoproteins and IHD. A 1mmol/L(39mg/dL) higher level of nonfasting remnant cholesterol was associated observationally with a 37%(95%CI:35-39%) higher C-reactive protein level, and causally with a 28%(10-48%) higher level. For LDL cholesterol, a 1mmol/L(39mg/dL) higher level was associated observationally with a 7%(6-7%) higher C-reactive protein level, but we found no causal association. Likewise, higher levels of C-reactive protein did not associate causally with elevated nonfasting remnant cholesterol or LDL cholesterol. Finally, the causal risk ratio for IHD for a 1mmol/L(39mg/dL) higher level was 3.3(95%CI: 2.1-5.2) for nonfasting remnant cholesterol and 1.8(1.5-2.2) for LDL cholesterol. The causal associations for remnant cholesterol were present even in those without diabetes mellitus and obesity.
Elevated nonfasting remnant cholesterol is causally associated with low-grade inflammation and with IHD, while elevated LDL cholesterol associated causally with IHD without inflammation.
[Show abstract][Hide abstract] ABSTRACT: Genome-wide association studies (GWAS) yielded significant advances in defining the genetic architecture of complex traits and disease. Still, a major hurdle of GWAS is narrowing down multiple genetic associations to a few causal variants for functional studies. This becomes critical in multi-phenotype GWAS where detection and interpretability of complex SNP(s)-trait(s) associations are complicated by complex Linkage Disequilibrium patterns between SNPs and correlation between traits. Here we propose a computationally efficient algorithm (GUESS) to explore complex genetic-association models and maximize genetic variant detection. We integrated our algorithm with a new Bayesian strategy for multi-phenotype analysis to identify the specific contribution of each SNP to different trait combinations and study genetic regulation of lipid metabolism in the Gutenberg Health Study (GHS). Despite the relatively small size of GHS (n = 3,175), when compared with the largest published meta-GWAS (n>100,000), GUESS recovered most of the major associations and was better at refining multi-trait associations than alternative methods. Amongst the new findings provided by GUESS, we revealed a strong association of SORT1 with TG-APOB and LIPC with TG-HDL phenotypic groups, which were overlooked in the larger meta-GWAS and not revealed by competing approaches, associations that we replicated in two independent cohorts. Moreover, we demonstrated the increased power of GUESS over alternative multi-phenotype approaches, both Bayesian and non-Bayesian, in a simulation study that mimics real-case scenarios. We showed that our parallel implementation based on Graphics Processing Units outperforms alternative multi-phenotype methods. Beyond multivariate modelling of multi-phenotypes, our Bayesian model employs a flexible hierarchical prior structure for genetic effects that adapts to any correlation structure of the predictors and increases the power to identify associated variants. This provides a powerful tool for the analysis of diverse genomic features, for instance including gene expression and exome sequencing data, where complex dependencies are present in the predictor space.