Quantifying the underestimation of relative risks from genome-wide association studies

PASTRY: achieving balanced power for detecting risk and protective minor alleles in meta-analysis of association studies with overlapping subjects

BACKGROUND

Meta-analysis is a statistical method that combines the results of multiple studies to increase statistical power. When multiple studies participating in a meta-analysis utilize the same public dataset as controls, the summary statistics from these studies become correlated. To solve this challenge, Lin and Sullivan proposed a method to provide an optimal test statistic adjusted for the correlation. This method quickly became the standard practice. However, we identified an unexpected power asymmetry phenomenon in this standard framework. This can lead to unbalanced power for detecting protective minor alleles and risk minor alleles.

RESULTS

We found that the power asymmetry of the current framework is mainly due to the errors in approximating the correlation term. We then developed a meta-analysis method based on an accurate correlation estimator, called PASTRY (A method to avoid Power ASymmeTRY). PASTRY outperformed the standard method on both simulated and real datasets in terms of the power symmetry.

CONCLUSIONS

Our findings suggest that PASTRY can help to alleviate the power asymmetry problem. PASTRY is available at https://github.com/hanlab-SNU/PASTRY .

Elucidating the Underlying Functional Mechanisms of Breast Cancer Susceptibility Through Post-GWAS Analyses

Genome-wide association studies (GWAS) have identified more than 170 single nucleotide polymorphisms (SNPs) associated with the susceptibility to breast cancer. Together, these SNPs explain 18% of the familial relative risk, which is estimated to be nearly half of the total familial breast cancer risk that is collectively explained by low-risk susceptibility alleles. An important aspect of this success has been the access to large sample sizes through collaborative efforts within the Breast Cancer Association Consortium (BCAC), but also collaborations between cancer association consortia. Despite these achievements, however, understanding of each variant's underlying mechanism and how these SNPs predispose women to breast cancer remains limited and represents a major challenge in the field, particularly since the vast majority of the GWAS-identified SNPs are located in non-coding regions of the genome and are merely tags for the causal variants. In recent years, fine-scale mapping studies followed by functional evaluation of putative causal variants have begun to elucidate the biological function of several GWAS-identified variants. In this review, we discuss the findings and lessons learned from these post-GWAS analyses of 22 risk loci. Identifying the true causal variants underlying breast cancer susceptibility and their function not only provides better estimates of the explained familial relative risk thereby improving polygenetic risk scores (PRSs), it also increases our understanding of the biological mechanisms responsible for causing susceptibility to breast cancer. This will facilitate the identification of further breast cancer risk alleles and the development of preventive medicine for those women at increased risk for developing the disease.

Common, germline genetic variations in the novel tumor suppressor BAP1 and risk of developing different types of cancer

BRCA1 associated protein-1 (BAP1) is a novel tumor suppressor that has recently been shown to be somatically mutated in several cancers. The BAP1 gene also carries rare germline mutations in families with a high incidence of several types of cancers, such as mesothelioma, uveal melanoma, lung adenocarcinoma, melanocytic neoplasms, and renal cell carcinoma. To test the hypothesis that common, germline genetic variants in BAP1 may also contribute to the risk of developing different types of cancer, we genotyped germline single nucleotide polymorphisms (SNPs) for BAP1 in a large population of patients with cancer, including 2,340 with colorectal cancer, 1,436 with bladder cancer, 3,313 with lung cancer, 1,325 with renal cell carcinoma, and 1,162 with esophageal cancer. We identified significant association of rs11708581 (P = 0.0034) and rs390802 (P = 0.015) with risk of renal cell carcinoma and rs12163565 (P = 0.038) with risk of lung cancer. Expression quantitative trait loci analysis in renal cell carcinoma using publicly available data from TCGA showed that the proxy SNPs for rs11708581 and rs390802 were negatively associated with the expression level of BAP1. Our study indicate that common germline genetic variants of BAP1 play a role in mediating the risk of developing renal cell carcinoma and lung cancer.

Case-Only Survival Analysis Reveals Unique Effects of Genotype, Sex, and Coronary Disease Severity on Survivorship

Survival bias may unduly impact genetic association with complex diseases; gene-specific survival effects may further complicate such investigations. Coronary artery disease (CAD) is a complex phenotype for which little is understood about gene-specific survival effects; yet, such information can offer insight into refining genetic associations, improving replications, and can provide candidate genes for both mortality risk and improved survivorship in CAD. Building on our previous work, the purpose of this current study was to: evaluate LSAMP SNP-specific hazards for all-cause mortality post-catheterization in a larger cohort of our CAD cases; and, perform additional replication in an independent dataset. We examined two LSAMP SNPs—rs1462845 and rs6788787—using CAD case-only Cox proportional hazards regression for additive genetic effects, censored on time-to-all-cause mortality or last follow-up among Caucasian subjects from the Catheterization Genetics Study (CATHGEN; n = 2,224) and the Intermountain Heart Collaborative Study (IMHC; n = 3,008). Only after controlling for age, sex, body mass index, histories of smoking, type 2 diabetes, hyperlipidemia and hypertension (HR = 1.11, 95%CI = 1.01–1.22, p = 0.032), rs1462845 conferred significantly increased hazards of all-cause mortality among CAD cases. Even after controlling for multiple covariates, but in only the primary cohort, rs6788787 conferred significantly improved survival (HR = 0.80, 95% CI = 0.69–0.92, p = 0.002). Post-hoc analyses further stratifying by sex and disease severity revealed replicated effects for rs1462845: even after adjusting for aforementioned covariates and coronary interventional procedures, males with severe burden of CAD had significantly amplified hazards of death with the minor variant of rs1462845 in both cohorts (HR = 1.29, 95% CI = 1.08–1.55, p = 0.00456; replication HR = 1.25, 95% CI = 1.05–1.49, p = 0.013). Kaplan-Meier curves revealed unique cohort-specific genotype effects on survival. Additional analyses demonstrated that the homozygous risk genotype (‘A/A’) fully explained the increased hazard in both cohorts. None of the post-hoc analyses in control subjects were significant for any model. This suggests that genetic effects of rs1462845 on survival are unique to CAD presence. This represents formal, replicated evidence of genetic contribution of rs1462845 to increased risk for all-cause mortality; the contribution is unique to CAD case status and specific to males with severe burden of CAD.

Imputation of KIR Types from SNP Variation Data

Large population studies of immune system genes are essential for characterizing their role in diseases, including autoimmune conditions. Of key interest are a group of genes encoding the killer cell immunoglobulin-like receptors (KIRs), which have known and hypothesized roles in autoimmune diseases, resistance to viruses, reproductive conditions, and cancer. These genes are highly polymorphic, which makes typing expensive and time consuming. Consequently, despite their importance, KIRs have been little studied in large cohorts. Statistical imputation methods developed for other complex loci (e.g., human leukocyte antigen [HLA]) on the basis of SNP data provide an inexpensive high-throughput alternative to direct laboratory typing of these loci and have enabled important findings and insights for many diseases. We present KIR^∗IMP, a method for imputation of KIR copy number. We show that KIR^∗IMP is highly accurate and thus allows the study of KIRs in large cohorts and enables detailed investigation of the role of KIRs in human disease.

Dissecting Allele Architecture of Early Onset IBD Using High-Density Genotyping

Background

The inflammatory bowel diseases (IBD) are common, complex disorders in which genetic and environmental factors are believed to interact leading to chronic inflammatory responses against the gut microbiota. Earlier genetic studies performed in mostly adult population of European descent identified 163 loci affecting IBD risk, but most have relatively modest effect sizes, and altogether explain only ~20% of the genetic susceptibility. Pediatric onset represents about 25% of overall incident cases in IBD, characterized by distinct disease physiology, course and risks. The goal of this study is to compare the allelic architecture of early onset IBD with adult onset in population of European descent.

Methods

We performed a fine mapping association study of early onset IBD using high-density Immunochip genotyping on 1008 pediatric-onset IBD cases (801 Crohn’s disease; 121 ulcerative colitis and 86 IBD undetermined) and 1633 healthy controls. Of the 158 SNP genotypes obtained (out of the 163 identified in adult onset), this study replicated 4% (5 SNPs out of 136) of the SNPs identified in the Crohn’s disease (CD) cases and 0.8% (1 SNP out of 128) in the ulcerative colitis (UC) cases. Replicated SNPs implicated the well known NOD2 and IL23R. The point estimate for the odds ratio (ORs) for NOD2 was above and outside the confidence intervals reported in adult onset. A polygenic liability score weakly predicted the age of onset for a larger collection of CD cases (p< 0.03, R²= 0.007), but not for the smaller number of UC cases.

Conclusions

The allelic architecture of common susceptibility variants for early onset IBD is similar to that of adult onset. This immunochip genotyping study failed to identify additional common variants that may explain the distinct phenotype that characterize early onset IBD. A comprehensive dissection of genetic loci is necessary to further characterize the genetic architecture of early onset IBD.

Fine mapping of genetic susceptibility loci for melanoma reveals a mixture of single variant and multiple variant regions

At least 17 genomic regions are established as harboring melanoma susceptibility variants, in most instances with genome-wide levels of significance and replication in independent samples. Based on genome-wide single nucleotide polymorphism (SNP) data augmented by imputation to the 1,000 Genomes reference panel, we have fine mapped these regions in over 5,000 individuals with melanoma (mainly from the GenoMEL consortium) and over 7,000 ethnically matched controls. A penalized regression approach was used to discover those SNP markers that most parsimoniously explain the observed association in each genomic region. For the majority of the regions, the signal is best explained by a single SNP, which sometimes, as in the tyrosinase region, is a known functional variant. However in five regions the explanation is more complex. At the CDKN2A locus, for example, there is strong evidence that not only multiple SNPs but also multiple genes are involved. Our results illustrate the variability in the biology underlying genome-wide susceptibility loci and make steps toward accounting for some of the "missing heritability."

Clinical genetic research 2: Genetic epidemiology of complex phenotypes

Genetic factors play a substantive role in the susceptibility to common diseases. Due to recent and rapid advancements in characterization of genetic variants and large-scale genotyping platforms, multiple genes and genetic variants have now been identified for common, complex diseases. The most efficient method for gene identification at present appears to be large-scale association-based studies, which integrate genetic and epidemiological principles. As the strategy for gene identification studies has shifted towards genetic association-based methods rather than traditional linkage analysis, epidemiological methods are increasingly being integrated into genetic investigations. Consequently, the disciplines of genetics and epidemiology, which historically have functioned separately, have been integrated into a discipline referred to as genetic epidemiology. In this chapter, we review methods for establishing the genetic burden of complex genetic disease, followed by methods for gene and/or genetic variant identification and when appropriate we highlight the epidemiological issues that guide these methods.

Cohen's h for detection of disease association with rare genetic variants

Background

The power of the genome wide association studies starts to go down when the minor allele frequency (MAF) is below 0.05. Here, we proposed the use of Cohen’s h in detecting disease associated rare variants. The variance stabilizing effect based on the arcsine square root transformation of MAFs to generate Cohen’s h contributed to the statistical power for rare variants analysis. We re-analyzed published datasets, one microarray and one sequencing based, and used simulation to compare the performance of Cohen’s h with the risk difference (RD) and odds ratio (OR).

Results

The analysis showed that the type 1 error rate of Cohen’s h was as expected and Cohen’s h and RD were both less biased and had higher power than OR. The advantage of Cohen’s h was more obvious when MAF was less than 0.01.

Conclusions

Cohen’s h can increase the power to find genetic association of rare variants and diseases, especially when MAF is less than 0.01.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-875) contains supplementary material, which is available to authorized users.

A note on the efficiencies of sampling strategies in two-stage Bayesian regional fine mapping of a quantitative trait

In focused studies designed to follow up associations detected in a genome-wide association study (GWAS), investigators can proceed to fine-map a genomic region by targeted sequencing or dense genotyping of all variants in the region, aiming to identify a functional sequence variant. For the analysis of a quantitative trait, we consider a Bayesian approach to fine-mapping study design that incorporates stratification according to a promising GWAS tag SNP in the same region. Improved cost-efficiency can be achieved when the fine-mapping phase incorporates a two-stage design, with identification of a smaller set of more promising variants in a subsample taken in stage 1, followed by their evaluation in an independent stage 2 subsample. To avoid the potential negative impact of genetic model misspecification on inference we incorporate genetic model selection based on posterior probabilities for each competing model. Our simulation study shows that, compared to simple random sampling that ignores genetic information from GWAS, tag-SNP-based stratified sample allocation methods reduce the number of variants continuing to stage 2 and are more likely to promote the functional sequence variant into confirmation studies.

Genome wide association studies in presence of misclassified binary responses

BACKGROUND

Misclassification has been shown to have a high prevalence in binary responses in both livestock and human populations. Leaving these errors uncorrected before analyses will have a negative impact on the overall goal of genome-wide association studies (GWAS) including reducing predictive power. A liability threshold model that contemplates misclassification was developed to assess the effects of mis-diagnostic errors on GWAS. Four simulated scenarios of case-control datasets were generated. Each dataset consisted of 2000 individuals and was analyzed with varying odds ratios of the influential SNPs and misclassification rates of 5% and 10%.

RESULTS

Analyses of binary responses subject to misclassification resulted in underestimation of influential SNPs and failed to estimate the true magnitude and direction of the effects. Once the misclassification algorithm was applied there was a 12% to 29% increase in accuracy, and a substantial reduction in bias. The proposed method was able to capture the majority of the most significant SNPs that were not identified in the analysis of the misclassified data. In fact, in one of the simulation scenarios, 33% of the influential SNPs were not identified using the misclassified data, compared with the analysis using the data without misclassification. However, using the proposed method, only 13% were not identified. Furthermore, the proposed method was able to identify with high probability a large portion of the truly misclassified observations.

CONCLUSIONS

The proposed model provides a statistical tool to correct or at least attenuate the negative effects of misclassified binary responses in GWAS. Across different levels of misclassification probability as well as odds ratios of significant SNPs, the model proved to be robust. In fact, SNP effects, and misclassification probability were accurately estimated and the truly misclassified observations were identified with high probabilities compared to non-misclassified responses. This study was limited to situations where the misclassification probability was assumed to be the same in cases and controls which is not always the case based on real human disease data. Thus, it is of interest to evaluate the performance of the proposed model in that situation which is the current focus of our research.

Quantifying missing heritability at known GWAS loci

Recent work has shown that much of the missing heritability of complex traits can be resolved by estimates of heritability explained by all genotyped SNPs. However, it is currently unknown how much heritability is missing due to poor tagging or additional causal variants at known GWAS loci. Here, we use variance components to quantify the heritability explained by all SNPs at known GWAS loci in nine diseases from WTCCC1 and WTCCC2. After accounting for expectation, we observed all SNPs at known GWAS loci to explain more heritability than GWAS-associated SNPs on average (). For some diseases, this increase was individually significant: for Multiple Sclerosis (MS) () and for Crohn's Disease (CD) (); all analyses of autoimmune diseases excluded the well-studied MHC region. Additionally, we found that GWAS loci from other related traits also explained significant heritability. The union of all autoimmune disease loci explained more MS heritability than known MS SNPs () and more CD heritability than known CD SNPs (), with an analogous increase for all autoimmune diseases analyzed. We also observed significant increases in an analysis of Rheumatoid Arthritis (RA) samples typed on ImmunoChip, with more heritability from all SNPs at GWAS loci () and more heritability from all autoimmune disease loci () compared to known RA SNPs (including those identified in this cohort). Our methods adjust for LD between SNPs, which can bias standard estimates of heritability from SNPs even if all causal variants are typed. By comparing adjusted estimates, we hypothesize that the genome-wide distribution of causal variants is enriched for low-frequency alleles, but that causal variants at known GWAS loci are skewed towards common alleles. These findings have important ramifications for fine-mapping study design and our understanding of complex disease architecture.

Heritable diseases have an unknown underlying “genetic architecture” that defines the distribution of effect-sizes for disease-causing mutations. Understanding this genetic architecture is an important first step in designing disease-mapping studies, and many theories have been developed on the nature of this distribution. Here, we evaluate the hypothesis that additional heritable variation lies at previously known associated loci but is not fully explained by the single most associated marker. We develop methods based on variance-components analysis to quantify this type of “local” heritability, demonstrating that standard strategies can be falsely inflated or deflated due to correlation between neighboring markers and propose a robust adjustment. In analysis of nine common diseases we find a significant average increase of local heritability, consistent with multiple common causal variants at an average locus. Intriguingly, for autoimmune diseases we also observe significant local heritability in loci not associated with the specific disease but with other autoimmune diseases, implying a highly correlated underlying disease architecture. These findings have important implications to the design of future studies and our general understanding of common disease.

Genome-wide association study in NSAID-induced acute urticaria/angioedema in Spanish and Han Chinese populations

AIM

Acute urticaria/angioedema (AUA) induced by cross-intolerance to NSAIDs is the most frequent clinical entity in hypersensitivity reactions to drugs. In this work, we conducted a genome-wide association study in Spanish and Han Chinese patients suffering from NSAID-induced AUA.

MATERIALS & METHODS

A whole-genome scan was performed on a total of 232 cases (112 Spanish and 120 Han Chinese) with NSAID-induced AUA and 225 unrelated controls (124 Spanish and 101 Han Chinese).

RESULTS

Although no polymorphism reached genome-wide significance, we obtained suggestive associations for three clusters in the Spanish group (RIMS1, BICC1 and RAD51L 1) and one region in the Han Chinese population (ABI3BP). Five regions showed suggestive associations after meta-analysis: HLF, RAD51L1, COL24A1, GalNAc-T13 and FBXL7. A majority of these genes are related to Ca(2+), cAMP and/or P53 signaling pathways.

CONCLUSION

The associations described were different from those related to the metabolism of arachidonic acid and could provide new mechanisms underlying NSAID-induced AUA.

Using genetic prediction from known complex disease Loci to guide the design of next-generation sequencing experiments

A central focus of complex disease genetics after genome-wide association studies (GWAS) is to identify low frequency and rare risk variants, which may account for an important fraction of disease heritability unexplained by GWAS. A profusion of studies using next-generation sequencing are seeking such risk alleles. We describe how already-known complex trait loci (largely from GWAS) can be used to guide the design of these new studies by selecting cases, controls, or families who are most likely to harbor undiscovered risk alleles. We show that genetic risk prediction can select unrelated cases from large cohorts who are enriched for unknown risk factors, or multiply-affected families that are more likely to harbor high-penetrance risk alleles. We derive the frequency of an undiscovered risk allele in selected cases and controls, and show how this relates to the variance explained by the risk score, the disease prevalence and the population frequency of the risk allele. We also describe a new method for informing the design of sequencing studies using genetic risk prediction in large partially-genotyped families using an extension of the Inside-Outside algorithm for inference on trees. We explore several study design scenarios using both simulated and real data, and show that in many cases genetic risk prediction can provide significant increases in power to detect low-frequency and rare risk alleles. The same approach can also be used to aid discovery of non-genetic risk factors, suggesting possible future utility of genetic risk prediction in conventional epidemiology. Software implementing the methods in this paper is available in the R package Mangrove.

Genome-wide association of body fat distribution in African ancestry populations suggests new loci

Central obesity, measured by waist circumference (WC) or waist-hip ratio (WHR), is a marker of body fat distribution. Although obesity disproportionately affects minority populations, few studies have conducted genome-wide association study (GWAS) of fat distribution among those of predominantly African ancestry (AA). We performed GWAS of WC and WHR, adjusted and unadjusted for BMI, in up to 33,591 and 27,350 AA individuals, respectively. We identified loci associated with fat distribution in AA individuals using meta-analyses of GWA results for WC and WHR (stage 1). Overall, 25 SNPs with single genomic control (GC)-corrected p-values<5.0 × 10(-6) were followed-up (stage 2) in AA with WC and with WHR. Additionally, we interrogated genomic regions of previously identified European ancestry (EA) WHR loci among AA. In joint analysis of association results including both Stage 1 and 2 cohorts, 2 SNPs demonstrated association, rs2075064 at LHX2, p = 2.24×10(-8) for WC-adjusted-for-BMI, and rs6931262 at RREB1, p = 2.48×10(-8) for WHR-adjusted-for-BMI. However, neither signal was genome-wide significant after double GC-correction (LHX2: p = 6.5 × 10(-8); RREB1: p = 5.7 × 10(-8)). Six of fourteen previously reported loci for waist in EA populations were significant (p<0.05 divided by the number of independent SNPs within the region) in AA studied here (TBX15-WARS2, GRB14, ADAMTS9, LY86, RSPO3, ITPR2-SSPN). Further, we observed associations with metabolic traits: rs13389219 at GRB14 associated with HDL-cholesterol, triglycerides, and fasting insulin, and rs13060013 at ADAMTS9 with HDL-cholesterol and fasting insulin. Finally, we observed nominal evidence for sexual dimorphism, with stronger results in AA women at the GRB14 locus (p for interaction = 0.02). In conclusion, we identified two suggestive loci associated with fat distribution in AA populations in addition to confirming 6 loci previously identified in populations of EA. These findings reinforce the concept that there are fat distribution loci that are independent of generalized adiposity.

Bringing genome-wide association findings into clinical use

Genome-wide association studies (GWASs) have been heralded as a major advance in biomedical discovery, having identified ~2,000 robust associations with complex diseases since 2005. Despite this success, they have met considerable scepticism regarding their clinical applicability; this scepticism arises from such aspects as the modest effect sizes of associated variants and their unclear functional consequences. There are, however, promising examples of GWAS findings that will or that may soon be translated into clinical care. These examples include variants identified through GWASs that provide strongly predictive or prognostic information or that have important pharmacological implications; these examples may illustrate promising approaches to wider clinical application.

Imputation-based meta-analysis of severe malaria in three African populations

Combining data from genome-wide association studies (GWAS) conducted at different locations, using genotype imputation and fixed-effects meta-analysis, has been a powerful approach for dissecting complex disease genetics in populations of European ancestry. Here we investigate the feasibility of applying the same approach in Africa, where genetic diversity, both within and between populations, is far more extensive. We analyse genome-wide data from approximately 5,000 individuals with severe malaria and 7,000 population controls from three different locations in Africa. Our results show that the standard approach is well powered to detect known malaria susceptibility loci when sample sizes are large, and that modern methods for association analysis can control the potential confounding effects of population structure. We show that pattern of association around the haemoglobin S allele differs substantially across populations due to differences in haplotype structure. Motivated by these observations we consider new approaches to association analysis that might prove valuable for multicentre GWAS in Africa: we relax the assumptions of SNP-based fixed effect analysis; we apply Bayesian approaches to allow for heterogeneity in the effect of an allele on risk across studies; and we introduce a region-based test to allow for heterogeneity in the location of causal alleles.

The minor allele of the missense polymorphism Ser251Pro in perilipin 2 (PLIN2) disrupts an α-helix, affects lipolysis, and is associated with reduced plasma triglyceride concentration in humans

Perilipin 2 (PLIN2) is the most abundant lipid droplet (LD)-associated protein in nonadipose tissue, and its expression correlates with intracellular lipid accumulation. Here we identified a missense polymorphism, Ser251Pro, that has major effect on protein structure and function, along with an influence on human plasma triglyceride concentration. The evolutionarily conserved Ser251Pro polymorphism was identified with the ClustalW program. Structure modeling using 3D-JigSaw and the Chimera package revealed that the Pro251 allele disrupts a predicted α-helix in PLIN2. Analyses of macrophages from individuals carrying Ser251Pro variants and human embryonic kidney 293 (HEK293) cells stably transfected with either of the alleles demonstrated that the Pro251 variant causes increased lipid accumulation and decreased lipolysis. Analysis of LD size distribution in stably transfected cells showed that the minor Pro251 allele resulted in an increased number of small LDs per cell and increased perilipin 3 protein expression levels as compared with cells carrying the major Ser251 allele. Genotyping of 2113 individuals indicated that the Pro251 variant is associated with decreased plasma triglyceride and very low-density lipoprotein concentrations. Altogether, these data provide the first evidence of a polymorphism in PLIN2 that affects PLIN2 function and may influence the development of metabolic and cardiovascular diseases.

Trans-ethnic fine-mapping of lipid loci identifies population-specific signals and allelic heterogeneity that increases the trait variance explained

Genome-wide association studies (GWAS) have identified ∼100 loci associated with blood lipid levels, but much of the trait heritability remains unexplained, and at most loci the identities of the trait-influencing variants remain unknown. We conducted a trans-ethnic fine-mapping study at 18, 22, and 18 GWAS loci on the Metabochip for their association with triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C), respectively, in individuals of African American (n = 6,832), East Asian (n = 9,449), and European (n = 10,829) ancestry. We aimed to identify the variants with strongest association at each locus, identify additional and population-specific signals, refine association signals, and assess the relative significance of previously described functional variants. Among the 58 loci, 33 exhibited evidence of association at P<1×10⁻⁴ in at least one ancestry group. Sequential conditional analyses revealed that ten, nine, and four loci in African Americans, Europeans, and East Asians, respectively, exhibited two or more signals. At these loci, accounting for all signals led to a 1.3- to 1.8-fold increase in the explained phenotypic variance compared to the strongest signals. Distinct signals across ancestry groups were identified at PCSK9 and APOA5. Trans-ethnic analyses narrowed the signals to smaller sets of variants at GCKR, PPP1R3B, ABO, LCAT, and ABCA1. Of 27 variants reported previously to have functional effects, 74% exhibited the strongest association at the respective signal. In conclusion, trans-ethnic high-density genotyping and analysis confirm the presence of allelic heterogeneity, allow the identification of population-specific variants, and limit the number of candidate SNPs for functional studies.

Lipid traits are heritable, but many of the DNA variants that influence lipid levels remain unknown. In a genomic region, more than one variant may affect gene expression or function, and the frequencies of these variants can differ across populations. Genotyping densely spaced variants in individuals with different ancestries may increase the chance of identifying variants that affect gene expression or function. We analyzed high-density genotyped variants for association with TG, HDL-C, and LDL-C in African Americans, East Asians, and Europeans. At several genomic regions, we provide evidence that two or more variants can influence lipid traits; across loci, these additional signals increase the proportion of trait variation that can be explained by genes. At some association signals shared across populations, combining data from individuals of different ancestries narrowed the set of likely functional variants. At PCSK9 and APOA5, the data suggest that different variants influence trait levels in different populations. Variants previously reported to alter gene expression or function frequently exhibited the strongest association at those signals. The multiple signals and population-specific characteristics of the loci described here may be shared by genetic loci for other complex traits.

From markers to molecular mechanisms: type 1 diabetes in the post-GWAS era

By the year 2000, a draft of the human genome sequence was completed. Millions of single-nucleotide polymorphisms (SNPs) had been deposited into public databases, and high throughput technologies were under development for SNP genotyping. At that time, it was predicted that large case control association studies would provide far better resolution and power than genome-wide linkage studies. Type 1 diabetes was one of the first phenotypes to be examined by genome-wide association studies (GWAS), and to date over 50 genomic regions have been associated with the disease. In general, the great majority of these loci individually contribute a relatively small degree of risk, and most loci lie outside of coding sequences. The identification of molecular mechanisms from these genomic data therefore remains a significant challenge. Here, we summarize genetic candidate, linkage, and association studies of type 1 diabetes and discuss a potential strategy to identify mechanisms of disease from genomic data.

Identifying breast cancer risk loci by global differential allele-specific expression (DASE) analysis in mammary epithelial transcriptome

BACKGROUND

The significant mortality associated with breast cancer (BCa) suggests a need to improve current research strategies to identify new genes that predispose women to breast cancer. Differential allele-specific expression (DASE) has been shown to contribute to phenotypic variables in humans and recently to the pathogenesis of cancer. We previously reported that nonsense-mediated mRNA decay (NMD) could lead to DASE of BRCA1/2, which is associated with elevated susceptibility to breast cancer. In addition to truncation mutations, multiple genetic and epigenetic factors can contribute to DASE, and we propose that DASE is a functional index for cis-acting regulatory variants and pathogenic mutations, and that global analysis of DASE in breast cancer precursor tissues can be used to identify novel causative alleles for breast cancer susceptibility.

RESULTS

To test our hypothesis, we employed the Illumina(®) Omni1-Quad BeadChip in paired genomic DNA (gDNA) and double-stranded cDNA (ds-cDNA) samples prepared from eight BCa patient-derived normal mammary epithelial lines (HMEC). We filtered original array data according to heterozygous genotype calls and calculated DASE values using the Log ratio of cDNA allele intensity, which was normalized to the corresponding gDNA. We developed two statistical methods, SNP- and gene-based approaches, which allowed us to identify a list of 60 candidate DASE loci (DASE ≥ 2.00, P ≤ 0.01, FDR ≤ 0.05) by both methods. Ingenuity Pathway Analysis of DASE loci revealed one major breast cancer-relevant interaction network, which includes two known cancer causative genes, ZNF331 (DASE = 2.31, P = 0.0018, FDR = 0.040) and USP6 (DASE = 4.80, P = 0.0013, FDR = 0.013), and a breast cancer causative gene, DMBT1 (DASE=2.03, P = 0.0017, FDR = 0.014). Sequence analysis of a 5' RACE product of DMBT1 demonstrated that rs2981745, a putative breast cancer risk locus, appears to be one of the causal variants leading to DASE in DMBT1.

CONCLUSIONS

Our study demonstrated for the first time that global DASE analysis is a powerful new approach to identify breast cancer risk allele(s).

Predicting signatures of "synthetic associations" and "natural associations" from empirical patterns of human genetic variation

Genome-wide association studies (GWAS) have in recent years discovered thousands of associated markers for hundreds of phenotypes. However, associated loci often only explain a relatively small fraction of heritability and the link between association and causality has yet to be uncovered for most loci. Rare causal variants have been suggested as one scenario that may partially explain these shortcomings. Specifically, Dickson et al. recently reported simulations of rare causal variants that lead to association signals of common, tag single nucleotide polymorphisms, dubbed "synthetic associations". However, an open question is what practical implications synthetic associations have for GWAS. Here, we explore the signatures exhibited by such "synthetic associations" and their implications based on patterns of genetic variation observed in human populations, thus accounting for human evolutionary history -a force disregarded in previous simulation studies. This is made possible by human population genetic data from HapMap 3 consisting of both resequencing and array-based genotyping data for the same set of individuals from multiple populations. We report that synthetic associations tend to be further away from the underlying risk alleles compared to "natural associations" (i.e. associations due to underlying common causal variants), but to a much lesser extent than previously predicted, with both the age and the effect size of the risk allele playing a part in this phenomenon. We find that while a synthetic association has a lower probability of capturing causal variants within its linkage disequilibrium block, sequencing around the associated variant need not extend substantially to have a high probability of capturing at least one causal variant. We also show that the minor allele frequency of synthetic associations is lower than of natural associations for most, but not all, loci that we explored. Finally, we find the variance in associated allele frequency to be a potential indicator of synthetic associations.

Elements of 'missing heritability'

PURPOSE OF REVIEW

To discuss the basis of 'missing heritability', which has emerged as an enigma in the post-genome-wide association studies (GWAS) era.

RECENT FINDINGS

Alleles identified through GWAS account for a relatively small fraction of heritability of the complex phenotypes. Accordingly, a significant part of heritability of the complex traits remains unaccounted for ('missing heritability'). Recent findings offer several explanations, including overestimation of heritability of the complex traits and underestimation of the effects of alleles identified through GWAS. In addition, yet-to-be identified common as well as rare alleles might in part explain the 'missing heritability'. Moreover, gene-gene (epistasis) and gene-environmental interactions might explain another fraction of heritability of complex traits. Moreover, transgenerational epigenetic changes, regulated in part by microRNAs, might also contribute to the 'missing heritability'.

SUMMARY

The new findings suggest a multifarious nature of the 'missing heritability'. The findings de-emphasize the focus on delineating the basis of 'missing heritability' and shift the focus to elucidation of the molecular mechanisms by which genomic and genetic factors govern the pathogenesis of the complex phenotypes.

A two-platform design for next generation genome-wide association studies

Genome-wide association studies (GWAS) have been successful in their search for common genetic variants associated with complex traits and diseases. With new advances in array technologies together with available genetic reference sets, the next generation of GWAS will extend the search for associations with uncommon SNPs (1% ≤ MAF ≤ 10%). Two possible approaches are genotyping all participants, a prohibitively expensive option for large GWAS, or using a combination of genotyping and imputation. Here, we consider a two platform method that genotypes all participants on a standard genotyping array, designed to identify common variants, and then supplements that data by genotyping only a small proportion of the participants on a platform that has higher coverage for uncommon SNPs. This subset of the study population is then included as part of the imputation reference set. To demonstrate the use of this two-platform design, we evaluate its potential efficiency using a newly available dataset containing 756 individuals genotyped on both the Illumina Human OmniExpress and Omni2.5 Quad. Although genotyping all individuals on the denser array would be ideal, we find that genotyping only 100 individuals on this array, in combination with imputation, leads to only a modest loss of power for detecting associations. However, the loss of power due to imputation can be more substantial if the relative risks for rare variants are significantly larger than those previously observed for common variants.

Resolving the variable genome and epigenome in human disease

The individual human genome and epigenome are being defined at unprecedented resolution by current advances in sequencing technologies with important implications for human disease. This review uses examples relevant to clinical practice to illustrate the functional consequences of genetic and epigenetic variation. The insights gained from genome-wide association studies are described together with current efforts to understand the role of rare variants in common disease, set in the context of recent successes in Mendelian traits through the application of whole exome sequencing. The application of functional genomics to interrogate the genome and epigenome, build up an integrated picture of the regulatory genomic landscape and inform disease association studies is discussed, together with the role of expression quantitative trait mapping and analysis of allele-specific gene expression.

Bayesian inference analyses of the polygenic architecture of rheumatoid arthritis

The genetic architectures of common, complex diseases are largely uncharacterized. We modeled the genetic architecture underlying genome-wide association study (GWAS) data for rheumatoid arthritis and developed a new method using polygenic risk-score analyses to infer the total liability-scale variance explained by associated GWAS SNPs. Using this method, we estimated that, together, thousands of SNPs from rheumatoid arthritis GWAS explain an additional 20% of disease risk (excluding known associated loci). We further tested this method on datasets for three additional diseases and obtained comparable estimates for celiac disease (43% excluding the major histocompatibility complex), myocardial infarction and coronary artery disease (48%) and type 2 diabetes (49%). Our results are consistent with simulated genetic models in which hundreds of associated loci harbor common causal variants and a smaller number of loci harbor multiple rare causal variants. These analyses suggest that GWAS will continue to be highly productive for the discovery of additional susceptibility loci for common diseases.

Dense genotyping identifies and localizes multiple common and rare variant association signals in celiac disease

Using variants from the 1000 Genomes Project pilot European CEU dataset and data from additional resequencing studies, we densely genotyped 183 non-HLA risk loci previously associated with immune-mediated diseases in 12,041 individuals with celiac disease (cases) and 12,228 controls. We identified 13 new celiac disease risk loci reaching genome-wide significance, bringing the number of known loci (including the HLA locus) to 40. We found multiple independent association signals at over one-third of these loci, a finding that is attributable to a combination of common, low-frequency and rare genetic variants. Compared to previously available data such as those from HapMap3, our dense genotyping in a large sample collection provided a higher resolution of the pattern of linkage disequilibrium and suggested localization of many signals to finer scale regions. In particular, 29 of the 54 fine-mapped signals seemed to be localized to single genes and, in some instances, to gene regulatory elements. Altogether, we define the complex genetic architecture of the risk regions of and refine the risk signals for celiac disease, providing the next step toward uncovering the causal mechanisms of the disease.

Estimating causal effects of genetic risk variants for breast cancer using marker data from bilateral and familial cases

BACKGROUND

Cases with a family history are enriched for genetic risk variants, and the power of association studies can be improved by selecting cases with a family history of disease. However, in recent genome-wide association scans utilizing familial sampling, the excess relative risk for familial cases is less than predicted when compared with unselected cases. This can be explained by incomplete linkage disequilibrium between the tested marker and the underlying causal variant.

METHODS

We show that the allele frequency and effect size of the underlying causal variant can be estimated by combining marker data from studies that ascertain cases based on different family histories. This allows us to learn about the genetic architecture of a complex trait, without having identified any causal variants. We consider several validated common marker alleles for breast cancer, using our own study of high risk, predominantly bilateral cases, cases preferentially selected to have at least two affected first- or second-degree relatives, and published estimates of relative risk from standard case-control studies.

RESULTS

To obtain realistic estimates and to accommodate some prior beliefs, we use Bayesian estimation to infer that the causal variants are probably common, with minor allele frequency >5%, and have small effects, with relative risk around 1.2.

CONCLUSION

These results strongly support the common disease common variant hypothesis for these specific loci associated with breast cancer.

IMPACT

Our results agree with recent assertions that synthetic associations of rare variants are unlikely to account for most associations seen in genome-wide studies.

Fine mapping of five loci associated with low-density lipoprotein cholesterol detects variants that double the explained heritability

Complex trait genome-wide association studies (GWAS) provide an efficient strategy for evaluating large numbers of common variants in large numbers of individuals and for identifying trait-associated variants. Nevertheless, GWAS often leave much of the trait heritability unexplained. We hypothesized that some of this unexplained heritability might be due to common and rare variants that reside in GWAS identified loci but lack appropriate proxies in modern genotyping arrays. To assess this hypothesis, we re-examined 7 genes (APOE, APOC1, APOC2, SORT1, LDLR, APOB, and PCSK9) in 5 loci associated with low-density lipoprotein cholesterol (LDL-C) in multiple GWAS. For each gene, we first catalogued genetic variation by re-sequencing 256 Sardinian individuals with extreme LDL-C values. Next, we genotyped variants identified by us and by the 1000 Genomes Project (totaling 3,277 SNPs) in 5,524 volunteers. We found that in one locus (PCSK9) the GWAS signal could be explained by a previously described low-frequency variant and that in three loci (PCSK9, APOE, and LDLR) there were additional variants independently associated with LDL-C, including a novel and rare LDLR variant that seems specific to Sardinians. Overall, this more detailed assessment of SNP variation in these loci increased estimates of the heritability of LDL-C accounted for by these genes from 3.1% to 6.5%. All association signals and the heritability estimates were successfully confirmed in a sample of ∼10,000 Finnish and Norwegian individuals. Our results thus suggest that focusing on variants accessible via GWAS can lead to clear underestimates of the trait heritability explained by a set of loci. Further, our results suggest that, as prelude to large-scale sequencing efforts, targeted re-sequencing efforts paired with large-scale genotyping will increase estimates of complex trait heritability explained by known loci.

Despite the striking success of genome-wide association studies in identifying genetic loci associated with common complex traits and diseases, much of the heritable risk for these traits and diseases remains unexplained. A higher resolution investigation of the genome through sequencing studies is expected to clarify the sources of this missing heritability. As a preview of what we might learn in these more detailed assessments of genetic variation, we used sequencing to identify potentially interesting variants in seven genes associated with low-density lipoprotein cholesterol (LDL-C) in 256 Sardinian individuals with extreme LDL-C levels, followed by large scale genotyping in 5,524 individuals, to examine newly discovered and previously described variants. We found that a combination of common and rare variants in these loci contributes to variation in LDL-C levels, and also that the initial estimate of the heritability explained by these loci doubled. Importantly, our results include a Sardinian-specific rare variant, highlighting the need for sequencing studies in isolated populations. Our results provide insights about what extensive whole-genome sequencing efforts are likely to reveal for the understanding of the genetic architecture of complex traits.

Identifying disease-associated SNP clusters via contiguous outlier detection

MOTIVATION

Although genome-wide association studies (GWAS) have identified many disease-susceptibility single-nucleotide polymorphisms (SNPs), these findings can only explain a small portion of genetic contributions to complex diseases, which is known as the missing heritability. A possible explanation is that genetic variants with small effects have not been detected. The chance is < 8 that a causal SNP will be directly genotyped. The effects of its neighboring SNPs may be too weak to be detected due to the effect decay caused by imperfect linkage disequilibrium. Moreover, it is still challenging to detect a causal SNP with a small effect even if it has been directly genotyped.

RESULTS

In order to increase the statistical power when detecting disease-associated SNPs with relatively small effects, we propose a method using neighborhood information. Since the disease-associated SNPs account for only a small fraction of the entire SNP set, we formulate this problem as Contiguous Outlier DEtection (CODE), which is a discrete optimization problem. In our formulation, we cast the disease-associated SNPs as outliers and further impose a spatial continuity constraint for outlier detection. We show that this optimization can be solved exactly using graph cuts. We also employ the stability selection strategy to control the false positive results caused by imperfect parameter tuning. We demonstrate its advantage in simulations and real experiments. In particular, the newly identified SNP clusters are replicable in two independent datasets.

AVAILABILITY

The software is available at: http://bioinformatics.ust.hk/CODE.zip.

CONTACT

eeyu@ust.hk

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

The genetics of asthma and allergic disease: a 21st century perspective

Asthma and allergy are common conditions with complex etiologies involving both genetic and environmental contributions. Recent genome-wide association studies (GWAS) and meta-analyses of GWAS have begun to shed light on both common and distinct pathways that contribute to asthma and allergic diseases. Associations with variation in genes encoding the epithelial cell-derived cytokines, interleukin-33 (IL-33) and thymic stromal lymphopoietin (TSLP), and the IL1RL1 gene encoding the IL-33 receptor, ST2, highlight the central roles for innate immune response pathways that promote the activation and differentiation of T-helper 2 cells in the pathogenesis of both asthma and allergic diseases. In contrast, variation at the 17q21 asthma locus, encoding the ORMDL3 and GSDML genes, is specifically associated with risk for childhood onset asthma. These and other genetic findings are providing a list of well-validated asthma and allergy susceptibility genes that are expanding our understanding of the common and unique biological pathways that are dysregulated in these related conditions. Ongoing studies will continue to broaden our understanding of asthma and allergy and unravel the mechanisms for the development of these complex traits.