Family-based association tests for genomewide association scans

Robust relationship inference in genome-wide association studies

MOTIVATION

Genome-wide association studies (GWASs) have been widely used to map loci contributing to variation in complex traits and risk of diseases in humans. Accurate specification of familial relationships is crucial for family-based GWAS, as well as in population-based GWAS with unknown (or unrecognized) family structure. The family structure in a GWAS should be routinely investigated using the SNP data prior to the analysis of population structure or phenotype. Existing algorithms for relationship inference have a major weakness of estimating allele frequencies at each SNP from the entire sample, under a strong assumption of homogeneous population structure. This assumption is often untenable.

RESULTS

Here, we present a rapid algorithm for relationship inference using high-throughput genotype data typical of GWAS that allows the presence of unknown population substructure. The relationship of any pair of individuals can be precisely inferred by robust estimation of their kinship coefficient, independent of sample composition or population structure (sample invariance). We present simulation experiments to demonstrate that the algorithm has sufficient power to provide reliable inference on millions of unrelated pairs and thousands of relative pairs (up to 3rd-degree relationships). Application of our robust algorithm to HapMap and GWAS datasets demonstrates that it performs properly even under extreme population stratification, while algorithms assuming a homogeneous population give systematically biased results. Our extremely efficient implementation performs relationship inference on millions of pairs of individuals in a matter of minutes, dozens of times faster than the most efficient existing algorithm known to us.

AVAILABILITY

Our robust relationship inference algorithm is implemented in a freely available software package, KING, available for download at http://people.virginia.edu/∼wc9c/KING.

A locus on chromosome 1p36 is associated with thyrotropin and thyroid function as identified by genome-wide association study

Thyroid hormones are key regulators of cellular growth, development, and metabolism, and thyroid disorders are a common cause of ill health in the community. Circulating concentrations of thyrotropin (TSH), thyroxine (T4) and triiodothyronine (T3) have a strong heritable component and are thought to be under polygenic control, but the genes responsible are mostly unknown. In order to identify genetic loci associated with these metabolic phenotypes, we performed a genome-wide association study of 2,120,505 SNPs in 2014 female twins from the TwinsUK study and found a significant association between rs10917469 on chromosome 1p36.13 and serum TSH (p = 3.2 × 10(-8)). The association of rs10917469 with serum TSH was replicated (p = 2.0 × 10(-4)) in an independent community-based sample of 1154 participants in the Busselton Health Study. This SNP is located near CAPZB, which might be a regulator of TSH secretion and thus of pituitary-thyroid axis function. Twenty-nine percent of white individuals carry the variant, and the difference in mean TSH concentrations between wild-type individuals and those homozygous for the minor G allele was 0.5 mU/l, which is likely to be clinically relevant. We also provide evidence of suggestive association (p < 5.0 × 10(-6)) of other SNPs with serum TSH, free T4, and free T3 concentrations, and these SNPs might be good targets for further studies. These results advance understanding of the genetic basis of pituitary-thyroid axis function and metabolic regulation.

New loci associated with central cornea thickness include COL5A1, AKAP13 and AVGR8

Central corneal thickness (CCT) is a highly heritable trait, which has been proposed to influence disorders of the anterior segment of the eye. A genome-wide association study (GWAS) of CCT was performed in 2269 individuals from three Croatian and one Scottish population. In the discovery set (1445 individuals), two genome-wide significant associations were identified for single nucleotide polymorphisms rs12447690 (β = 0.23 SD, P = 4.4 × 10(-9)) and rs1536482 (β = 0.22 SD, P = 7.1 × 10(-8)) for which the closest candidate genes (although ≥90 kb away) were zinc finger 469 (ZNF469) on 16q24.2 and collagen 5 alpha 1 (COL5A1) on 9q34.2, respectively. Only the ZNF469 association was confirmed in our replication set (824 individuals, P = 8.0 × 10(-4)) but COL5A1 remained a suggestive association in the combined sample (β = 0.16 SD, P = 1.1 × 10(-6)). Following a larger meta-analysis including recently published CCT GWAS summary data, COL5A1 was genome-wide significant (β = 0.13 SD, P = 5.1 × 10(-8)), together with two additional novel loci. The second new locus (defined by rs1034200) was 5 kb from the AVGR8 gene, encoding a putative transcription factor with typical ZNF and KRAB domains, in chromosomal region 13q12.11 (β = 0.14 SD, P = 3.5 × 10(-9)). The third new locus (rs6496932), on 15q25.3 (β = 0.13, P = 1.4 × 10(-8)), was within a wide linkage disequilibrium block extending into the 5' end of the AKAP13 gene, encoding a scaffold protein concerned with signal transduction from the cell surface. These associations offer mechanistic insights into the regulation of CCT and offer new candidate genes for susceptibility to common disorders in which CCT has been implicated, including primary open-angle glaucoma and keratoconus.

A genome-wide association study of Cloninger's temperament scales: implications for the evolutionary genetics of personality

Variation in personality traits is 30-60% attributed to genetic influences. Attempts to unravel these genetic influences at the molecular level have, so far, been inconclusive. We performed the first genome-wide association study of Cloninger's temperament scales in a sample of 5117 individuals, in order to identify common genetic variants underlying variation in personality. Participants' scores on Harm Avoidance, Novelty Seeking, Reward Dependence, and Persistence were tested for association with 1,252,387 genetic markers. We also performed gene-based association tests and biological pathway analyses. No genetic variants that significantly contribute to personality variation were identified, while our sample provides over 90% power to detect variants that explain only 1% of the trait variance. This indicates that individual common genetic variants of this size or greater do not contribute to personality trait variation, which has important implications regarding the genetic architecture of personality and the evolutionary mechanisms by which heritable variation is maintained.

A genome-wide association study of self-rated health

Self-rated health questions have been proven to be a highly reliable and valid measure of overall health as measured by other indicators in many population groups. It also has been shown to be a very good predictor of mortality, chronic or severe diseases, and the need for services, and is positively correlated with clinical assessments. Genetic factors have been estimated to account for 25-64% of the variance in the liability of self-rated health. The aim of the present study was to identify Single Nucleotide Polymorphisms (SNPs) underlying the heritability of self-rated health by conducting a genome-wide association analysis in a large sample of 6,706 Australian individuals aged 18-92. No genome wide significant SNPs associated with self-rated health could be identified, indicating that self-rated health may be influenced by a large number of SNPs with very small effect size. A very large sample will be needed to identify these SNPs.

The perception of quinine taste intensity is associated with common genetic variants in a bitter receptor cluster on chromosome 12

The perceived taste intensities of quinine HCl, caffeine, sucrose octaacetate (SOA) and propylthiouracil (PROP) solutions were examined in 1457 twins and their siblings. Previous heritability modeling of these bitter stimuli indicated a common genetic factor for quinine, caffeine and SOA (22–28%), as well as separate specific genetic factors for PROP (72%) and quinine (15%). To identify the genes involved, we performed a genome-wide association study with the same sample as the modeling analysis, genotyped for approximately 610 000 single-nucleotide polymorphisms (SNPs). For caffeine and SOA, no SNP association reached a genome-wide statistical criterion. For PROP, the peak association was within TAS2R38 (rs713598, A49P, P = 1.6 × 10⁻¹⁰⁴), which accounted for 45.9% of the trait variance. For quinine, the peak association was centered in a region that contains bitter receptor as well as salivary protein genes and explained 5.8% of the trait variance (TAS2R19, rs10772420, R299C, P = 1.8 × 10⁻¹⁵). We confirmed this association in a replication sample of twins of similar ancestry (P = 0.00001). The specific genetic factor for the perceived intensity of PROP was identified as the gene previously implicated in this trait (TAS2R38). For quinine, one or more bitter receptor or salivary proline-rich protein genes on chromosome 12 have alleles which affect its perception but tight linkage among very similar genes precludes the identification of a single causal genetic variant.

Common genetic determinants of vitamin D insufficiency: a genome-wide association study

BACKGROUND

Vitamin D is crucial for maintenance of musculoskeletal health, and might also have a role in extraskeletal tissues. Determinants of circulating 25-hydroxyvitamin D concentrations include sun exposure and diet, but high heritability suggests that genetic factors could also play a part. We aimed to identify common genetic variants affecting vitamin D concentrations and risk of insufficiency.

METHODS

We undertook a genome-wide association study of 25-hydroxyvitamin D concentrations in 33 996 individuals of European descent from 15 cohorts. Five epidemiological cohorts were designated as discovery cohorts (n=16 125), five as in-silico replication cohorts (n=9367), and five as de-novo replication cohorts (n=8504). 25-hydroxyvitamin D concentrations were measured by radioimmunoassay, chemiluminescent assay, ELISA, or mass spectrometry. Vitamin D insufficiency was defined as concentrations lower than 75 nmol/L or 50 nmol/L. We combined results of genome-wide analyses across cohorts using Z-score-weighted meta-analysis. Genotype scores were constructed for confirmed variants.

FINDINGS

Variants at three loci reached genome-wide significance in discovery cohorts for association with 25-hydroxyvitamin D concentrations, and were confirmed in replication cohorts: 4p12 (overall p=1.9x10(-109) for rs2282679, in GC); 11q12 (p=2.1x10(-27) for rs12785878, near DHCR7); and 11p15 (p=3.3x10(-20) for rs10741657, near CYP2R1). Variants at an additional locus (20q13, CYP24A1) were genome-wide significant in the pooled sample (p=6.0x10(-10) for rs6013897). Participants with a genotype score (combining the three confirmed variants) in the highest quartile were at increased risk of having 25-hydroxyvitamin D concentrations lower than 75 nmol/L (OR 2.47, 95% CI 2.20-2.78, p=2.3x10(-48)) or lower than 50 nmol/L (1.92, 1.70-2.16, p=1.0x10(-26)) compared with those in the lowest quartile.

INTERPRETATION

Variants near genes involved in cholesterol synthesis, hydroxylation, and vitamin D transport affect vitamin D status. Genetic variation at these loci identifies individuals who have substantially raised risk of vitamin D insufficiency.

FUNDING

Full funding sources listed at end of paper (see Acknowledgments).

IRF4 variants have age-specific effects on nevus count and predispose to melanoma

High melanocytic nevus count is a strong predictor of melanoma risk. A GWAS of nevus count in Australian adolescent twins identified an association of nevus count with the interferon regulatory factor 4 gene (IRF4 [p = 6 x 10(-9)]). There was a strong genotype-by-age interaction, which was replicated in independent UK samples of adolescents and adults. The rs12203592(*)T allele was associated with high nevus counts and high freckling scores in adolescents, but with low nevus counts and high freckling scores in adults. The rs12203592(*)T increased counts of flat (compound and junctional) nevi in Australian adolescent twins, but decreased counts of raised (intradermal) nevi. In combined analysis of melanoma case-control data from Australia, the UK, and Sweden, the rs12203592(*)C allele was associated with melanoma (odds ratio [OR] 1.15, p = 4 x 10(-3)), most significantly on the trunk (OR = 1.33, p = 2.5 x 10(-5)). The melanoma association was corroborated in a GWAS performed by the GenoMEL consortium for an adjacent SNP, rs872071 (rs872071(*)T: OR 1.14, p = 0.0035; excluding Australian, the UK, and Swedish samples typed at rs12203592: OR 1.08, p = 0.08).

A genome-wide association study of neuroticism in a population-based sample

Neuroticism is a moderately heritable personality trait considered to be a risk factor for developing major depression, anxiety disorders and dementia. We performed a genome-wide association study in 2,235 participants drawn from a population-based study of neuroticism, making this the largest association study for neuroticism to date. Neuroticism was measured by the Eysenck Personality Questionnaire. After Quality Control, we analysed 430,000 autosomal SNPs together with an additional 1.2 million SNPs imputed with high quality from the Hap Map CEU samples. We found a very small effect of population stratification, corrected using one principal component, and some cryptic kinship that required no correction. NKAIN2 showed suggestive evidence of association with neuroticism as a main effect (p<10⁻⁶) and GPC6 showed suggestive evidence for interaction with age (p≈10⁻⁷). We found support for one previously-reported association (PDE4D), but failed to replicate other recent reports. These results suggest common SNP variation does not strongly influence neuroticism. Our study was powered to detect almost all SNPs explaining at least 2% of heritability, and so our results effectively exclude the existence of loci having a major effect on neuroticism.

Genetic variation of DKK3 may modify renal disease severity in ADPKD

Significant variation in the course of autosomal dominant polycystic kidney disease ( ADPKD) within families suggests the presence of effect modifiers. Recent studies of the variation within families harboring PKD1 mutations indicate that genetic background may account for 32 to 42% of the variance in estimated GFR (eGFR) before ESRD and 43 to 78% of the variance in age at ESRD onset, but the genetic modifiers are unknown. Here, we conducted a high-throughput single-nucleotide polymorphism (SNP) genotyping association study of 173 biological candidate genes in 794 white patients from 227 families with PKD1. We analyzed two primary outcomes: (1) eGFR and (2) time to ESRD (renal survival). For both outcomes, we used multidimensional scaling to correct for population structure and generalized estimating equations to account for the relatedness among individuals within the same family. We found suggestive associations between each of 12 SNPs and at least one of the renal outcomes. We genotyped these SNPs in a second set of 472 white patients from 229 families with PKD1 and performed a joint analysis on both cohorts. Three SNPs continued to show suggestive/significant association with eGFR at the Dickkopf 3 (DKK3) gene locus; no SNPs significantly associated with renal survival. DKK3 antagonizes Wnt/beta-catenin signaling, which may modulate renal cyst growth. Pending replication, our study suggests that genetic variation of DKK3 may modify severity of ADPKD resulting from PKD1 mutations.

A genome-wide association study of optic disc parameters

The optic nerve head is involved in many ophthalmic disorders, including common diseases such as myopia and open-angle glaucoma. Two of the most important parameters are the size of the optic disc area and the vertical cup-disc ratio (VCDR). Both are highly heritable but genetically largely undetermined. We performed a meta-analysis of genome-wide association (GWA) data to identify genetic variants associated with optic disc area and VCDR. The gene discovery included 7,360 unrelated individuals from the population-based Rotterdam Study I and Rotterdam Study II cohorts. These cohorts revealed two genome-wide significant loci for optic disc area, rs1192415 on chromosome 1p22 (p = 6.72×10⁻¹⁹) within 117 kb of the CDC7 gene and rs1900004 on chromosome 10q21.3-q22.1 (p = 2.67×10⁻³³) within 10 kb of the ATOH7 gene. They revealed two genome-wide significant loci for VCDR, rs1063192 on chromosome 9p21 (p = 6.15×10⁻¹¹) in the CDKN2B gene and rs10483727 on chromosome 14q22.3-q23 (p = 2.93×10⁻¹⁰) within 40 kbp of the SIX1 gene. Findings were replicated in two independent Dutch cohorts (Rotterdam Study III and Erasmus Rucphen Family study; N = 3,612), and the TwinsUK cohort (N = 843). Meta-analysis with the replication cohorts confirmed the four loci and revealed a third locus at 16q12.1 associated with optic disc area, and four other loci at 11q13, 13q13, 17q23 (borderline significant), and 22q12.1 for VCDR. ATOH7 was also associated with VCDR independent of optic disc area. Three of the loci were marginally associated with open-angle glaucoma. The protein pathways in which the loci of optic disc area are involved overlap with those identified for VCDR, suggesting a common genetic origin.

Morphologic characteristics of the optic nerve head are involved in many ophthalmic diseases. Its size, called the optic disc area, is an important measure and has been associated with e.g. myopia and open-angle glaucoma (OAG). Another important and clinical parameter of the optic disc is the vertical cup-disc ratio (VCDR). Although studies have shown a high heritability of optic disc area and VCDR, its genetic determinants are still undetermined. We therefore conducted a genome-wide association (GWA) study on these quantitative traits, using data of over 11,000 Caucasian participants, and related the findings to myopia and OAG. We found evidence for association of three loci with optic disc area: CDC7/TGFBR3 region, ATOH7, and SALL1; and six with VCDR: CDKN2B, SIX1, SCYL1, CHEK2, ATOH7, and DCLK1; and additionally one borderline significant locus: BCAS3. None of the loci could be related to myopia. There was marginal evidence for association of ATOH7, CDKN2B, and SIX1 with OAG, which remains to be confirmed. The present study reveals new insights into the physiological development of the optic nerve and may shed light on the pathophysiological protein pathways leading to (neuro-) ophthalmologic diseases such as OAG.

Genome-wide association scan for five major dimensions of personality

Personality traits are summarized by five broad dimensions with pervasive influences on major life outcomes, strong links to psychiatric disorders and clear heritable components. To identify genetic variants associated with each of the five dimensions of personality we performed a genome-wide association (GWA) scan of 3972 individuals from a genetically isolated population within Sardinia, Italy. On the basis of the analyses of 362 129 single-nucleotide polymorphisms we found several strong signals within or near genes previously implicated in psychiatric disorders. They include the association of neuroticism with SNAP25 (rs362584, P=5 x 10(-5)), extraversion with BDNF and two cadherin genes (CDH13 and CDH23; Ps<5 x 10(-5)), openness with CNTNAP2 (rs10251794, P=3 x 10(-5)), agreeableness with CLOCK (rs6832769, P=9 x 10(-6)) and conscientiousness with DYRK1A (rs2835731, P=3 x 10(-5)). Effect sizes were small (less than 1% of variance), and most failed to replicate in the follow-up independent samples (N up to 3903), though the association between agreeableness and CLOCK was supported in two of three replication samples (overall P=2 x 10(-5)). We infer that a large number of loci may influence personality traits and disorders, requiring larger sample sizes for the GWA approach to confidently identify associated genetic variants.

Association of matrix metalloproteinase gene polymorphisms with refractive error in Amish and Ashkenazi families

PURPOSE

Matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) are involved in scleral extracellular matrix remodeling and have shown differential expression in experimental myopia. The genetic association of refractive error and polymorphisms in MMP and TIMP genes in Old Order Amish (AMISH) and Ashkenazi Jewish (ASHK) families was investigated.

METHODS

Individuals from 55 AMISH and 63 ASHK families participated in the study. Ascertainment was designed to enrich the families for myopia; the mean spherical equivalent (MSE) refractive error (SD) was -1.61 (2.72) D in the AMISH, and -3.56 (3.32) D in the ASHK. One hundred forty-six common haplotype tagging SNPs covering 14 MMP and 4 TIMP genes were genotyped in 358 AMISH and 535 ASHK participants. Association analyses of MSE and the spherical component of refraction (SPH) were performed separately for the AMISH and the ASHK. Bonferroni-corrected significance thresholds and local false discovery rates were used to account for multiple testing.

RESULTS

After they were filtered for quality-control, 127 SNPs were included in the analyses. No polymorphisms showed statistically significant association to refraction in the ASHK (minimum P = 0.0132). In AMISH, two SNPs showed evidence of association with refractive phenotypes: rs1939008 (P = 0.00016 for SPH); and rs9928731 (P = 0.00026 for SPH). These markers were each estimated to explain <5% of the variance of SPH in the AMISH sample.

CONCLUSIONS

Statistically significant genetic associations of ocular refraction to polymorphisms near MMP1 and within MMP2 were identified in the AMISH but not among the ASHK families. The results suggest that the MMP1 and MMP2 genes are involved in refractive variation in the AMISH. Genetic and/or environmental heterogeneity most likely contribute to differences in association results between ethnic groups.

Linkage and genome-wide association analysis of obesity-related phenotypes: association of weight with the MGAT1 gene

As major risk-factors for diabetes and cardiovascular diseases, the genetic contribution to obesity-related traits has been of interest for decades. Recently, a limited number of common genetic variants, which have replicated in different populations, have been identified. One approach to increase the statistical power in genetic mapping studies is to focus on populations with increased levels of linkage disequilibrium (LD) and reduced genetic diversity. We have performed joint linkage and genome-wide association analyses for weight and BMI in 3,448 (linkage) and 3,925 (association) partly overlapping healthy individuals from five European populations. A total of four chromosomal regions (two for weight and two for BMI) showed suggestive linkage (lod >2.69) either in one of the populations or in the joint data. At the genome-wide level (nominal P < 1.6 x 10(-7), Bonferroni-adjusted P < 0.05) one single-nucleotide polymorphism (SNP) (rs12517906) (nominal P = 7.3 x 10(-8)) was associated with weight, whereas none with BMI. The SNP associated with weight is located close to MGAT1. The monoacylglycerol acyltransferase (MGAT) enzyme family is known to be involved in dietary fat absorption. There was no overlap between the linkage regions and the associated SNPs. Our results show that genetic effects influencing weight and BMI are shared across diverse European populations, even though some of these populations have experienced recent population bottlenecks and/or been affected by genetic drift. The analysis enabled us to identify a new candidate gene, MGAT1, associated with weight in women.

BDNF Val66Met is associated with introversion and interacts with 5-HTTLPR to influence neuroticism

Brain-derived neurotrophic factor (BDNF) regulates synaptic plasticity and neurotransmission, and has been linked to neuroticism, a major risk factor for psychiatric disorders. A recent genome-wide association (GWA) scan, however, found the BDNF Val66Met polymorphism (rs6265) associated with extraversion but not with neuroticism. In this study, we examine the links between BDNF and personality traits, assessed using the Revised NEO Personality Inventory (NEO-PI-R), in a sample from SardiNIA (n=1560) and the Baltimore Longitudinal Study of Aging (BLSA; n=1131). Consistent with GWA results, we found that BDNF Met carriers were more introverted. By contrast, in both samples and in a meta-analysis inclusive of published data (n=15251), we found no evidence for a main effect of BDNF Val66Met on neuroticism. Finally, on the basis of recent reports of an epistatic effect between BDNF and the serotonin transporter, we explored a Val66Met x 5-HTTLPR interaction in a larger SardiNIA sample (n=2333). We found that 5-HTTLPR LL carriers scored lower on neuroticism in the presence of the BDNF Val variant, but scored higher on neuroticism in the presence of the BDNF Met variant. Our findings support the association between the BDNF Met variant and introversion and suggest that BDNF interacts with the serotonin transporter gene to influence neuroticism.

ProbABEL package for genome-wide association analysis of imputed data

Background

Over the last few years, genome-wide association (GWA) studies became a tool of choice for the identification of loci associated with complex traits. Currently, imputed single nucleotide polymorphisms (SNP) data are frequently used in GWA analyzes. Correct analysis of imputed data calls for the implementation of specific methods which take genotype imputation uncertainty into account.

Results

We developed the ProbABEL software package for the analysis of genome-wide imputed SNP data and quantitative, binary, and time-till-event outcomes under linear, logistic, and Cox proportional hazards models, respectively. For quantitative traits, the package also implements a fast two-step mixed model-based score test for association in samples with differential relationships, facilitating analysis in family-based studies, studies performed in human genetically isolated populations and outbred animal populations.

Conclusions

ProbABEL package provides fast efficient way to analyze imputed data in genome-wide context and will facilitate future identification of complex trait loci.

Multistage analysis strategies for genome-wide association studies: summary of group 3 contributions to Genetic Analysis Workshop 16

This contribution summarizes the work done by six independent teams of investigators to identify the genetic and non-genetic variants that work together or independently to predispose to disease. The theme addressed in these studies is multistage strategies in the context of genome-wide association studies (GWAS). The work performed comes from Group 3 of the Genetic Analysis Workshop 16 held in St. Louis, Missouri in September 2008. These six studies represent a diversity of multistage methods of which five are applied to the North American Rheumatoid Arthritis Consortium rheumatoid arthritis case-control data, and one method is applied to the low-density lipoprotein phenotype in the Framingham Heart Study simulated data. In the first stage of analyses, the majority of studies used a variety of screening techniques to reduce the noise of single-nucleotide polymorphisms purportedly not involved in the phenotype of interest. Three studies analyzed the data using penalized regression models, either LASSO or the elastic net. The main result was a reconfirmation of the involvement of variants in the HLA region on chromosome 6 with rheumatoid arthritis. The hope is that the intense computational methods highlighted in this group of papers will become useful tools in future GWAS.

A meta-analysis of genome-wide data from five European isolates reveals an association of COL22A1, SYT1, and GABRR2 with serum creatinine level

BACKGROUND

Serum creatinine (S CR) is the most important biomarker for a quick and non-invasive assessment of kidney function in population-based surveys. A substantial proportion of the inter-individual variability in S CR level is explicable by genetic factors.

METHODS

We performed a meta-analysis of genome-wide association studies of S CR undertaken in five population isolates ('discovery cohorts'), all of which are part of the European Special Population Network (EUROSPAN) project. Genes showing the strongest evidence for an association with SCR (candidate loci) were replicated in two additional population-based samples ('replication cohorts').

RESULTS

After the discovery meta-analysis, 29 loci were selected for replication. Association between SCR level and polymorphisms in the collagen type XXII alpha 1 (COL22A1) gene, on chromosome 8, and in the synaptotagmin-1 (SYT1) gene, on chromosome 12, were successfully replicated in the replication cohorts (p value = 1.0 x 10(-6) and 1.7 x 10(-4), respectively). Evidence of association was also found for polymorphisms in a locus including the gamma-aminobutyric acid receptor rho-2 (GABRR2) gene and the ubiquitin-conjugating enzyme E2-J1 (UBE2J1) gene (replication p value = 3.6 x 10(-3)). Previously reported findings, associating glomerular filtration rate with SNPs in the uromodulin (UMOD) gene and in the schroom family member 3 (SCHROOM3) gene were also replicated.

CONCLUSIONS

While confirming earlier results, our study provides new insights in the understanding of the genetic basis of serum creatinine regulatory processes. In particular, the association with the genes SYT1 and GABRR2 corroborate previous findings that highlighted a possible role of the neurotransmitters GABAA receptors in the regulation of the glomerular basement membrane and a possible interaction between GABAA receptors and synaptotagmin-I at the podocyte level.

Variance component model to account for sample structure in genome-wide association studies

Although genome-wide association studies (GWASs) have identified numerous loci associated with complex traits, imprecise modeling of the genetic relatedness within study samples may cause substantial inflation of test statistics and possibly spurious associations. Variance component approaches, such as efficient mixed-model association (EMMA), can correct for a wide range of sample structures by explicitly accounting for pairwise relatedness between individuals, using high-density markers to model the phenotype distribution; but such approaches are computationally impractical. We report here a variance component approach implemented in publicly available software, EMMA eXpedited (EMMAX), that reduces the computational time for analyzing large GWAS data sets from years to hours. We apply this method to two human GWAS data sets, performing association analysis for ten quantitative traits from the Northern Finland Birth Cohort and seven common diseases from the Wellcome Trust Case Control Consortium. We find that EMMAX outperforms both principal component analysis and genomic control in correcting for sample structure.

Genetic association tests: a method for the joint analysis of family and case-control data

With the trend in molecular epidemiology towards both genome-wide association studies and complex modelling, the need for large sample sizes to detect small effects and to allow for the estimation of many parameters within a model continues to increase. Unfortunately, most methods of association analysis have been restricted to either a family-based or a case-control design, resulting in the lack of synthesis of data from multiple studies. Transmission disequilibrium-type methods for detecting linkage disequilibrium from family data were developed as an effective way of preventing the detection of association due to population stratification. Because these methods condition on parental genotype, however, they have precluded the joint analysis of family and case-control data, although methods for case-control data may not protect against population stratification and do not allow for familial correlations. We present here an extension of a family-based association analysis method for continuous traits that will simultaneously test for, and if necessary control for, population stratification. We further extend this method to analyse binary traits (and therefore family and case-control data together) and accurately to estimate genetic effects in the population, even when using an ascertained family sample. Finally, we present the power of this binary extension for both family-only and joint family and case-control data, and demonstrate the accuracy of the association parameter and variance components in an ascertained family sample.

Magnitude of stratification in human populations and impacts on genome wide association studies

Genome-wide association studies (GWAS) may be biased by population stratification (PS). We conducted empirical quantification of the magnitude of PS among human populations and its impact on GWAS. Liver tissues were collected from 979, 59 and 49 Caucasian Americans (CA), African Americans (AA) and Hispanic Americans (HA), respectively, and genotyped using Illumina650Y (Ilmn650Y) arrays. RNA was also isolated and hybridized to Agilent whole-genome gene expression arrays. We propose a new method (i.e., hgdp-eigen) for detecting PS by projecting genotype vectors for each sample to the eigenvector space defined by the Human Genetic Diversity Panel (HGDP). Further, we conducted GWAS to map expression quantitative trait loci (eQTL) for the approximately 40,000 liver gene expression traits monitored by the Agilent arrays. HGDP-eigen performed similarly to the conventional self-eigen methods in capturing PS. However, leveraging the HGDP offered a significant advantage in revealing the origins, directions and magnitude of PS. Adjusting for eigenvectors had minor impacts on eQTL detection rates in CA. In contrast, for AA and HA, adjustment dramatically reduced association findings. At an FDR = 10%, we identified 65 eQTLs in AA with the unadjusted analysis, but only 18 eQTLs after the eigenvector adjustment. Strikingly, 55 out of the 65 unadjusted AA eQTLs were validated in CA, indicating that the adjustment procedure significantly reduced GWAS power. A number of the 55 AA eQTLs validated in CA overlapped with published disease associated SNPs. For example, rs646776 and rs10903129 have previously been associated with lipid levels and coronary heart disease risk, however, the rs10903129 eQTL was missed in the eigenvector adjusted analysis.

Genome-wide association study of PR interval

The electrocardiographic PR interval reflects atrial and atrioventricular nodal conduction, disturbances of which increase risk of atrial fibrillation (AF). To identify underlying common genetic variation, we meta-analyzed genome-wide association results for PR interval from seven community-based studies of European-ancestry individuals in the CHARGE consortium: AGES, ARIC, CHS, FHS, KORA, Rotterdam Study, and SardiNIA (N=28,517). Statistically significant loci (P<5×10^-8) were tested for association with AF (N=5,741 cases). We identified nine loci associated with PR interval. At chromosome 3p22.2, we observed two independent associations in voltage gated sodium channel genes SCN10A and SCN5A, while six loci were near cardiac developmental genes CAV1/CAV2, NKX2-5 (CSX1), SOX5, WNT11, MEIS1, and TBX5/TBX3. Another signal was at ARHGAP24, a locus without known relevance to the heart. Five of the nine loci, SCN5A, SCN10A, NKX2-5, CAV1/CAV2, and SOX5, were also associated with AF (P<0.0056). Common genetic variation, particularly in ion channel and developmental genes, contributes significantly to atrial and atrioventricular conduction and to AF risk.

Modeling of environmental effects in genome-wide association studies identifies SLC2A2 and HP as novel loci influencing serum cholesterol levels

Genome-wide association studies (GWAS) have identified 38 larger genetic regions affecting classical blood lipid levels without adjusting for important environmental influences. We modeled diet and physical activity in a GWAS in order to identify novel loci affecting total cholesterol, LDL cholesterol, HDL cholesterol, and triglyceride levels. The Swedish (SE) EUROSPAN cohort (N(SE) = 656) was screened for candidate genes and the non-Swedish (NS) EUROSPAN cohorts (N(NS) = 3,282) were used for replication. In total, 3 SNPs were associated in the Swedish sample and were replicated in the non-Swedish cohorts. While SNP rs1532624 was a replication of the previously published association between CETP and HDL cholesterol, the other two were novel findings. For the latter SNPs, the p-value for association was substantially improved by inclusion of environmental covariates: SNP rs5400 (p(SE,unadjusted) = 3.6 x 10(-5), p(SE,adjusted) = 2.2 x 10(-6), p(NS,unadjusted) = 0.047) in the SLC2A2 (Glucose transporter type 2) and rs2000999 (p(SE,unadjusted) = 1.1 x 10(-3), p(SE,adjusted) = 3.8 x 10(-4), p(NS,unadjusted) = 0.035) in the HP gene (Haptoglobin-related protein precursor). Both showed evidence of association with total cholesterol. These results demonstrate that inclusion of important environmental factors in the analysis model can reveal new genetic susceptibility loci.

Two-stage testing strategies for genome-wide association studies in family-based designs

The analysis of genome-wide association studies (GWAS) poses statistical hurdles that have to be handled efficiently in order for the study to be successful. The two largest impediments in the analysis phase of the study are the multiple comparisons problem and maintaining robustness against confounding due to population admixture and stratification. For quantitative traits in family-based designs, Van Steen (1) proposed a two-stage testing strategy that can be considered a hybrid approach between family-based and population-based analysis. By including the population-based component into the family-based analysis, the Van Steen algorithm maximizes the statistical power, while at the same time, maintains the original robustness of family-based association tests (FBATs) (2-4). The Van Steen approach consists of two statistically independent steps, a screening step and a testing step. For all genotyped single nucleotide polymorphisms (SNPs), the screening step examines the evidence for association at a population-based level. Based on support for a potential genetic association from the screening step, the SNPs are prioritized for testing in the next step, where they are analyzed with a FBAT (3). By exploiting population-based information in the screening step that is not utilized in family-based association testing step, the two steps are statistically independent. Therefore, the use of the population-based data for the purposes of screening does not bias the FBAT statistic calculated in the testing step. Depending on the trait type and the ascertainment conditions, Van Steen-type testing strategies can achieve statistical power levels that are comparable to those of population-based studies with the same number of probands. In this chapter, we review the original Van Steen algorithm, its numerous extensions, and discuss its advantages and disadvantages.

Quantitative trait loci for CD4:CD8 lymphocyte ratio are associated with risk of type 1 diabetes and HIV-1 immune control

Abnormal expansion or depletion of particular lymphocyte subsets is associated with clinical manifestations such as HIV progression to AIDS and autoimmune disease. We sought to identify genetic predictors of lymphocyte levels and reasoned that these may play a role in immune-related diseases. We tested 2.3 million variants for association with five lymphocyte subsets, measured in 2538 individuals from the general population, including CD4+ T cells, CD8+ T cells, CD56+ natural killer (NK) cells, and the derived measure CD4:CD8 ratio. We identified two regions of strong association. The first was located in the major histocompatibility complex (MHC), with multiple SNPs strongly associated with CD4:CD8 ratio (rs2524054, p = 2.1 x 10(-28)). The second region was centered within a cluster of genes from the Schlafen family and was associated with NK cell levels (rs1838149, p = 6.1 x 10(-14)). The MHC association with CD4:CD8 replicated convincingly (p = 1.4 x 10(-9)) in an independent panel of 988 individuals. Conditional analyses indicate that there are two major independent quantitative trait loci (QTL) in the MHC region that regulate CD4:CD8 ratio: one is located in the class I cluster and influences CD8 levels, whereas the second is located in the class II cluster and regulates CD4 levels. Jointly, both QTL explained 8% of the variance in CD4:CD8 ratio. The class I variants are also strongly associated with durable host control of HIV, and class II variants are associated with type-1 diabetes, suggesting that genetic variation at the MHC may predispose one to immune-related diseases partly through disregulation of T cell homeostasis.

Comparison between single-marker analysis using Merlin and multi-marker analysis using LASSO for Framingham simulated data

We compared family-based single-marker association analysis using Merlin and multi-marker analysis using LASSO (least absolute shrinkage and selection operator) for the low-density lipoprotein phenotype at the first visit for all 200 replicates of the Genetic Analysis Workshop 16 Framingham simulated data sets. Using "answers," we selected single-nucleotide polymorphisms (SNPs) on chromosome 22 for comparison of results between single-marker and multi-marker analyses. For the major causal SNP rs2294207 on chromosome 22, both single-marker and multi-marker analyses provided similar results, indicating the importance of this SNP. For the 12 polygenic SNPs on the same chromosome, both single-marker and multi-marker analyses failed to provide statistically significant associations, indicating that their effects were too weak to be detected by either method. The main difference between the two methods was that for the 14 SNPs near the causal SNPs, p-values from Merlin were the next smallest, whereas LASSO often excluded these non-causal neighboring SNPs entirely from the first 10,000 models.

Genome-wide discovery of maternal effect variants

Many phenotypes may be influenced by the prenatal environment of the mother and/or maternal care, and these maternal effects may have a heritable component. We have implemented in the computer program SOLAR a variance components-based method for detecting indirect effects of maternal genotype on offspring phenotype. Of six phenotypes measured in three generations of the Framingham Heart Study, height showed the strongest evidence (P = 0.02) of maternal effect. We conducted a genome-wide association analysis for height, testing both the direct effect of the focal individual's genotype and the indirect effect of the maternal genotype. Offspring height showed suggestive evidence of association with maternal genotype for two single-nucleotide polymorphisms in the trafficking protein particle complex 9 gene TRAPPC9 (NIBP), which plays a role in neuronal NF-κB signalling. This work establishes a methodological framework for identifying genetic variants that may influence the contribution of the maternal environment to offspring phenotypes.

STK39 polymorphisms and blood pressure: an association study in British Caucasians and assessment of cis-acting influences on gene expression

Background

Blood pressure (BP) has significant heritability, but the genes responsible remain largely unknown. Single nucleotide polymorphisms (SNPs) at the STK39 locus were recently associated with hypertension by genome-wide association in an Amish population; in vitro data from transient transfection experiments using reporter constructs suggested that altered STK39 expression might mediate the effect. However, other large studies have not implicated STK39 in hypertension. We determined whether reported SNPs influenced STK39 expression in vivo, or were associated with BP in a large British Caucasian cohort.

Methods

1372 members of 247 Caucasian families ascertained through a hypertensive proband were genotyped for reported risk variants in STK39 (rs6749447, rs3754777, rs35929607) using Sequenom technology. MERLIN software was used for family-based association testing. Cis-acting influences on expression were assessed in vivo using allelic expression ratios in cDNA from peripheral blood cells in 35 South African individuals heterozygous for a transcribed SNP in STK39 (rs1061471) and quantified by mass spectrometry (Sequenom).

Results

No significant association was seen between the SNPs tested and systolic or diastolic BP in clinic or ambulatory measurements (all p > 0.05). The tested SNPs were all associated with allelic expression differences in peripheral blood cells (p < 0.05), with the most significant association for the intronic SNP rs6749447 (P = 9.9 × 10^-4). In individuals who were heterozygous for this SNP, on average the G allele showed 13% overexpression compared to the T allele.

Conclusions

STK39 expression is modified by polymorphisms acting in cis and the typed SNPs are associated with allelic expression of this gene, but there is no evidence for an association with BP in a British Caucasian cohort.

Variants of the serotonin transporter gene and NEO-PI-R Neuroticism: No association in the BLSA and SardiNIA samples

The polymorphism in the serotonin transporter gene promoter region (5-HTTLPR) is by far the most studied variant hypothesized to influence Neuroticism-related personality traits. The results of previous studies have been mixed and appear moderated by the personality questionnaire used. Studies that used the TCI to assess Harm Avoidance or the EPQ to assess Neuroticism have found no association with the 5-HTTLPR. However, studies that used the NEO-PI-R or related instruments (NEO-PI, NEO-FFI) to measure Neuroticism have found some evidence of association. This study examines the association of variants in the serotonin transporter gene in a sample from a genetically isolated population within Sardinia (Italy) that is several times larger than previous samples that used the NEO-PI-R (N = 3,913). The association was also tested in a sample (N = 548) from the Baltimore Longitudinal Study of Aging (BLSA), in which repeated NEO-PI-R assessments were obtained. In the SardiNIA sample, we found no significant association of the 5-HTTLPR genotypes with Neuroticism or its facets (Anxiety, Angry-Hostility, Depression, Self-Consciousness, Impulsiveness, and Vulnerability). In the BLSA sample, we found lower scores on Neuroticism traits for the heterozygous group, which is inconsistent with previous studies. We also examined eight SNPs in the SardiNIA (N = 3,972) and nine SNPs in the BLSA (N = 1,182) that map within or near the serotonin transporter gene (SLC6A4), and found no association. Along with other large studies that used different phenotypic measures and found no association, this study substantially increases the evidence against a link between 5-HTT variants and Neuroticism-related traits.

Family-based bivariate association tests for quantitative traits

The availability of a large number of dense SNPs, high-throughput genotyping and computation methods promotes the application of family-based association tests. While most of the current family-based analyses focus only on individual traits, joint analyses of correlated traits can extract more information and potentially improve the statistical power. However, current TDT-based methods are low-powered. Here, we develop a method for tests of association for bivariate quantitative traits in families. In particular, we correct for population stratification by the use of an integration of principal component analysis and TDT. A score test statistic in the variance-components model is proposed. Extensive simulation studies indicate that the proposed method not only outperforms approaches limited to individual traits when pleiotropic effect is present, but also surpasses the power of two popular bivariate association tests termed FBAT-GEE and FBAT-PC, respectively, while correcting for population stratification. When applied to the GAW16 datasets, the proposed method successfully identifies at the genome-wide level the two SNPs that present pleiotropic effects to HDL and TG traits.

Maximizing the Power of Genome-Wide Association Studies: A Novel Class of Powerful Family-Based Association Tests

For genome-wide association studies in family-based designs, a new, universally applicable approach is proposed. Using a modified Liptak's method, we combine the p-value of the family-based association test (FBAT) statistic with the p-value for the Van Steen-statistic. The Van Steen-statistic is independent of the FBAT-statistic and utilizes information that is ignored by traditional FBAT-approaches. The new test statistic takes advantages of all available information about the genetic association, while, by virtue of its design, it achieves complete robustness against confounding due to population stratification. The approach is suitable for the analysis of almost any trait type for which FBATs are available, e.g. binary, continuous, time to-onset, multivariate, etc. The efficiency and the validity of the new approach depend on the specification of a nuisance/tuning parameter and the weight parameters in the modified Liptak's method. For different trait types and ascertainment conditions, we discuss general guidelines for the optimal specification of the tuning parameter and the weight parameters. Our simulation experiments and an application to an Alzheimer study show the validity and the efficiency of the new method, which achieves power levels that are comparable to those of population-based approaches.

Association and interaction analyses of eight genes under asthma linkage peaks

BACKGROUND

Linkage studies have implicated the 2q33, 9p21, 11q13 and 20q13 regions in the regulation of allergic disease. The aim of this study was to test genetic variants in candidate genes from these regions for association with specific asthma traits.

METHODS

Ninety-five single nucleotide polymorphisms (SNP) located in eight genes (CD28, CTLA4, ICOS, ADAM23, ADAMTSL1, MS4A2, CDH26 and HRH3) were genotyped in >5000 individuals from Australian (n = 1162), Dutch (n = 99) and Danish (n = 303) families. Traits tested included doctor-diagnosed asthma, atopy, airway obstruction, total serum immunoglobulin (Ig) E levels and eosinophilia. Association was tested using both multivariate and univariate methods, with gene-wide thresholds for significance determined through simulation. Gene-by-gene and gene-by-environment analyses were also performed.

RESULTS

There was no overall evidence for association with seven of the eight genes tested when considering all genetic variation assayed in each gene. The exception was MS4A2 on chromosome 11q13, which showed weak evidence for association with IgE (gene-wide P < 0.05, rs502581). There were no significant gene-by-gene or gene-by-environment interaction effects after accounting for the number of tests performed.

CONCLUSIONS

The individual variants genotyped in the 2q33, 9p21 and 20q13 regions do not explain a large fraction of the variation in the quantitative traits tested or have a major impact on asthma or atopy risk. Our results are consistent with a weak effect of MS4A2 polymorphisms on the variation of total IgE levels.

Sequence variants in three loci influence monocyte counts and erythrocyte volume

Blood cells participate in vital physiological processes, and their numbers are tightly regulated so that homeostasis is maintained. Disruption of key regulatory mechanisms underlies many blood-related Mendelian diseases but also contributes to more common disorders, including atherosclerosis. We searched for quantitative trait loci (QTL) for hematology traits through a whole-genome association study, because these could provide new insights into both hemopoeitic and disease mechanisms. We tested 1.8 million variants for association with 13 hematology traits measured in 6015 individuals from the Australian and Dutch populations. These traits included hemoglobin composition, platelet counts, and red blood cell and white blood cell indices. We identified three regions of strong association that, to our knowledge, have not been previously reported in the literature. The first was located in an intergenic region of chromosome 9q31 near LPAR1, explaining 1.5% of the variation in monocyte counts (best SNP rs7023923, p=8.9x10(-14)). The second locus was located on chromosome 6p21 and associated with mean cell erythrocyte volume (rs12661667, p=1.2x10(-9), 0.7% variance explained) in a region that spanned five genes, including CCND3, a member of the D-cyclin gene family that is involved in hematopoietic stem cell expansion. The third region was also associated with erythrocyte volume and was located in an intergenic region on chromosome 6q24 (rs592423, p=5.3x10(-9), 0.6% variance explained). All three loci replicated in an independent panel of 1543 individuals (p values=0.001, 9.9x10(-5), and 7x10(-5), respectively). The identification of these QTL provides new opportunities for furthering our understanding of the mechanisms regulating hemopoietic cell fate.

On the analysis of genome-wide association studies in family-based designs: a universal, robust analysis approach and an application to four genome-wide association studies

For genome-wide association studies in family-based designs, we propose a new, universally applicable approach. The new test statistic exploits all available information about the association, while, by virtue of its design, it maintains the same robustness against population admixture as traditional family-based approaches that are based exclusively on the within-family information. The approach is suitable for the analysis of almost any trait type, e.g. binary, continuous, time-to-onset, multivariate, etc., and combinations of those. We use simulation studies to verify all theoretically derived properties of the approach, estimate its power, and compare it with other standard approaches. We illustrate the practical implications of the new analysis method by an application to a lung-function phenotype, forced expiratory volume in one second (FEV1) in 4 genome-wide association studies.

Heritability and demographic analyses in the large isolated population of Val Borbera suggest advantages in mapping complex traits genes

BACKGROUND

Isolated populations are a useful resource for mapping complex traits due to shared stable environment, reduced genetic complexity and extended Linkage Disequilibrium (LD) compared to the general population. Here we describe a large genetic isolate from the North West Apennines, the mountain range that runs through Italy from the North West Alps to the South.

METHODOLOGY/PRINCIPAL FINDINGS

The study involved 1,803 people living in 7 villages of the upper Borbera Valley. For this large population cohort, data from genealogy reconstruction, medical questionnaires, blood, anthropometric and bone status QUS parameters were evaluated. Demographic and epidemiological analyses indicated a substantial genetic component contributing to each trait variation as well as overlapping genetic determinants and family clustering for some traits.

CONCLUSIONS/SIGNIFICANCE

The data provide evidence for significant heritability of medical relevant traits that will be important in mapping quantitative traits. We suggest that this population isolate is suitable to identify rare variants associated with complex phenotypes that may be difficult to study in larger but more heterogeneous populations.

Genetic determinants of circulating sphingolipid concentrations in European populations

Sphingolipids have essential roles as structural components of cell membranes and in cell signalling, and disruption of their metabolism causes several diseases, with diverse neurological, psychiatric, and metabolic consequences. Increasingly, variants within a few of the genes that encode enzymes involved in sphingolipid metabolism are being associated with complex disease phenotypes. Direct experimental evidence supports a role of specific sphingolipid species in several common complex chronic disease processes including atherosclerotic plaque formation, myocardial infarction (MI), cardiomyopathy, pancreatic beta-cell failure, insulin resistance, and type 2 diabetes mellitus. Therefore, sphingolipids represent novel and important intermediate phenotypes for genetic analysis, yet little is known about the major genetic variants that influence their circulating levels in the general population. We performed a genome-wide association study (GWAS) between 318,237 single-nucleotide polymorphisms (SNPs) and levels of circulating sphingomyelin (SM), dihydrosphingomyelin (Dih-SM), ceramide (Cer), and glucosylceramide (GluCer) single lipid species (33 traits); and 43 matched metabolite ratios measured in 4,400 subjects from five diverse European populations. Associated variants (32) in five genomic regions were identified with genome-wide significant corrected p-values ranging down to 9.08x10(-66). The strongest associations were observed in or near 7 genes functionally involved in ceramide biosynthesis and trafficking: SPTLC3, LASS4, SGPP1, ATP10D, and FADS1-3. Variants in 3 loci (ATP10D, FADS3, and SPTLC3) associate with MI in a series of three German MI studies. An additional 70 variants across 23 candidate genes involved in sphingolipid-metabolizing pathways also demonstrate association (p = 10(-4) or less). Circulating concentrations of several key components in sphingolipid metabolism are thus under strong genetic control, and variants in these loci can be tested for a role in the development of common cardiovascular, metabolic, neurological, and psychiatric diseases.

A generalized family-based association test for dichotomous traits

Recent advances in genotyping technology make it possible to utilize large-scale association analysis for disease-gene mapping. Powerful and robust family-based association methods are crucial for successful gene mapping. We propose a family-based association method, the generalized disequilibrium test (GDT), in which the genotype differences of all discordant relative pairs are utilized in assessing association within a family. The improvement of the GDT over existing methods is threefold: (1) information beyond first-degree relatives is incorporated efficiently, yielding substantial gains in power in comparison to existing tests; (2) the GDT statistic is implemented via a robust technique that does not rely on large sample theory, resulting in further power gains, especially at high levels of significance; and (3) covariates and weights based on family size are incorporated. Advantages of the GDT over existing methods are demonstrated by extensive computer simulations and by application to recently published large-scale genome-wide linkage data from the Type 1 Diabetes Genetics Consortium (T1DGC). In our simulations, the GDT consistently outperforms other tests for a common disease and frequently outperforms other tests for a rare disease; the power improvement is > 13% in 6 out of 8 extended pedigree scenarios. All of the six strongest associations identified by the GDT have been reported by other studies, whereas only three or four of these associations can be identified by existing methods. For the T1D association at gene UBASH3A, the GDT resulted in a genome-wide significance (p = 4.3 x 10(-6)), much stronger than the published significance (p = 10(-4)).

Tests of association for quantitative traits in nuclear families using principal components to correct for population stratification

Traditional transmission disequilibrium test (TDT) based methods for genetic association analyses are robust to population stratification at the cost of a substantial loss of power. We here describe a novel method for family-based association studies that corrects for population stratification with the use of an extension of principal component analysis (PCA). Specifically, we adopt PCA on unrelated parents in each family. We then infer principal components for children from those for their parents through a TDT-like strategy. Two test statistics within the variance-components model are proposed for association tests. Simulation results show that the proposed tests have correct type I error rates regardless of population stratification, and have greatly improved power over two popular TDT-based methods: QTDT and FBAT. The application to the Genetic Analysis Workshop 16 (GAW16) data sets attests to the feasibility of the proposed method.

Using new tools to define the genetic underpinnings of risky traits associated with coronary artery disease: the SardiNIA study

Genomewide association studies are increasingly being applied to search for novel genes that might underlie cardiovascular diseases. In this article, we briefly review the principles that underlie modern genetic analyses and provide several illustrations from the SardiNIA study of genomewide association studies for cardiovascular risk factor traits.

Genome-wide association studies and the genetic dissection of complex traits

The availability of affordable high throughput technology for parallel genotyping has opened the field of genetics to genome-wide association studies (GWAS), and in the last few years hundreds of articles reporting results of GWAS for a variety of heritable traits have been published. What do these results tell us? Although GWAS have discovered a few hundred reproducible associations, this number is underwhelming in relation to the huge amount of data produced, and challenges the conjecture that common variants may be the genetic causes of common diseases. We argue that the massive amount of genetic data that result from these studies remains largely unexplored and unexploited because of the challenge of mining and modeling enormous data sets, the difficulty of using nontraditional computational techniques and the focus of accepted statistical analyses on controlling the false positive rate rather than limiting the false negative rate. In this article, we will review the common approach to analysis of GWAS data and then discuss options to learn more from these data. We will use examples from our ongoing studies of sickle cell anemia and also GWAS in multigenic traits.

Genome-wide association study identifies variants at 9p21 and 22q13 associated with development of cutaneous nevi

A high melanocytic nevi count is the strongest known risk factor for cutaneous melanoma. We conducted a genome-wide association study for nevus count using 297,108 SNPs in 1,524 twins, with validation in an independent cohort of 4,107 individuals. We identified strongly associated variants in MTAP, a gene adjacent to the familial melanoma susceptibility locus CDKN2A on 9p21 (rs4636294, combined P = 3.4 x 10(-15)), as well as in PLA2G6 on 22q13.1 (rs2284063, combined P = 3.4 x 10(-8)). In addition, variants in these two loci showed association with melanoma risk in 3,131 melanoma cases from two independent studies, including rs10757257 at 9p21, combined P = 3.4 x 10(-8), OR = 1.23 (95% CI = 1.15-1.30) and rs132985 at 22q13.1, combined P = 2.6 x 10(-7), OR = 1.23 (95% CI = 1.15-1.30). This provides the first report of common variants associated to nevus number and demonstrates association of these variants with melanoma susceptibility.

Bivariate association analyses for the mixture of continuous and binary traits with the use of extended generalized estimating equations

Genome-wide association (GWA) study is becoming a powerful tool in deciphering genetic basis of complex human diseases/traits. Currently, the univariate analysis is the most commonly used method to identify genes associated with a certain disease/phenotype under study. A major limitation with the univariate analysis is that it may not make use of the information of multiple correlated phenotypes, which are usually measured and collected in practical studies. The multivariate analysis has proven to be a powerful approach in linkage studies of complex diseases/traits, but it has received little attention in GWA. In this study, we aim to develop a bivariate analytical method for GWA study, which can be used for a complex situation in which continuous trait and a binary trait are measured under study. Based on the modified extended generalized estimating equation (EGEE) method we proposed herein, we assessed the performance of our bivariate analyses through extensive simulations as well as real data analyses. In the study, to develop an EGEE approach for bivariate genetic analyses, we combined two different generalized linear models corresponding to phenotypic variables using a seemingly unrelated regression model. The simulation results demonstrated that our EGEE-based bivariate analytical method outperforms univariate analyses in increasing statistical power under a variety of simulation scenarios. Notably, EGEE-based bivariate analyses have consistent advantages over univariate analyses whether or not there exists a phenotypic correlation between the two traits. Our study has practical importance, as one can always use multivariate analyses as a screening tool when multiple phenotypes are available, without extra costs of statistical power and false-positive rate. Analyses on empirical GWA data further affirm the advantages of our bivariate analytical method.

Genome-wide association study of anthropometric traits in Korcula Island, Croatia

AIM

To identify genetic variants underlying six anthropometric traits: body height, body weight, body mass index, brachial circumference, waist circumference, and hip circumference, using a genome-wide association study.

METHODS

The study was carried out in the isolated population of the island of Korcula, Croatia, with 898 adult examinees who participated in the larger DNA-based genetic epidemiological study in 2007. Anthropometric measurements followed standard internationally accepted procedures. Examinees were genotyped using HumanHap 370CNV chip by Illumina, with a genome-wide scan containing 316730 single nucleotide polymorphisms (SNP).

RESULTS

A total of 11 SNPs were associated with the investigated traits at the level of P<10(-5), with one SNP (rs7792939 in gene zinc finger protein 498, ZNF498) associated with body weight, hip circumference, and brachial circumference (P=3.59-5.73 x 10(-6)), and another one (rs157350 in gene delta-sarcoglycan, SGCD) with both brachial and hip circumference (P=3.70-6.08 x 10(-6). Variants in CRIM1, a gene regulating delivery of bone morphogenetic proteins to the cell surface, and ITGA1, involved in the regulation of mesenchymal stem cell proliferation and cartilage production, were also associated with brachial circumference (P=7.82 and 9.68 x 10(-6), respectively) and represent interesting functional candidates. Other associations involved those between genes SEZ6L2 and MAX and waist circumference, XTP6 and brachial circumference, and AMPA1/GRIA1 and height.

CONCLUSION

Although the study was underpowered for the reported associations to reach formal threshold of genome-wide significance under the assumption of independent multiple testing, the consistency of association between the 2 variants and a set of anthropometric traits makes CRIM1 and ITGA1 highly interesting for further replication and functional follow-up. Increased linkage disequilibrium between the used markers in an isolated population makes the formal significance threshold overly stringent, and changed allele frequencies in isolate population may contribute to identifying variants that would not be easily identified in large outbred populations.

Genome-wide association study of biochemical traits in Korcula Island, Croatia

AIM

To identify genetic variants underlying biochemical traits--total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, uric acid, albumin, and fibrinogen, in a genome-wide association study in an isolated population where rare variants of larger effect may be more easily identified.

METHODS

The study included 944 adult inhabitants of the island of Korcula, as a part of larger DNA-based genetic epidemiological study in 2007. Biochemical measurements were performed in a single laboratory with stringent internal and external quality control procedures. Examinees were genotyped using Human Hap370CNV chip by Illumina, with a genome-wide scan containing 346027 single nucleotide polymorphisms (SNP).

RESULTS

A total of 31 SNPs were associated with 7 investigated traits at the level of P<1.00 x 10(-5). Nine of SNPs implicated the role of SLC2A9 in uric acid regulation (P=4.10 x 10(-6)-2.58 x 10(-12)), as previously found in other populations. All 22 remaining associations fell into the P=1.00 x 10(-5)-1.00 x 10(-6) significance range. One of them replicated the association between cholesteryl ester transfer protein (CETP) and HDL, and 7 associations were more than 100 kilobases away from the closest known gene. Nearby SNPs, rs4767631 and rs10444502, in gene kinase suppressor of ras 2 (KSR2) on chromosome 12 were associated with LDL cholesterol levels, and rs10444502 in the same gene with total cholesterol levels. Similarly, rs2839619 in gene PBX/knotted 1 homeobox 1 (PKNOX1) on chromosome 21 was associated with total and LDL cholesterol levels. The remaining 9 findings implied possible associations between phosphatidylethanolamine N-methyltransferase (PEMT) gene and total cholesterol; USP46, RAP1GDS1, and ZCCHC16 genes and triglycerides; BCAT1 and SLC14A2 genes and albumin; and NR3C2, GRIK2, and PCSK2 genes and fibrinogen.

CONCLUSION

Although this study was underpowered for most of the reported associations to reach formal threshold of genome-wide significance under the assumption of independent multiple testing, replications of previous findings and consistency of association between the identified variants and more than one studied trait make such findings interesting for further functional follow-up studies. Changed allele frequencies in isolate population may contribute to identifying variants that would not be easily identified in much larger samples in outbred populations.

A kinship-based modification of the armitage trend test to address hidden population structure and small differential genotyping errors

BACKGROUND/AIMS

We propose a modification of the well-known Armitage trend test to address the problems associated with hidden population structure and hidden relatedness in genome-wide case-control association studies.

METHODS

The new test adopts beneficial traits from three existing testing strategies: the principal components, mixed model, and genomic control while avoiding some of their disadvantageous characteristics, such as the tendency of the principal components method to over-correct in certain situations or the failure of the genomic control approach to reorder the adjusted tests based on their degree of alignment with the underlying hidden structure. The new procedure is based on Gauss-Markov estimators derived from a straightforward linear model with an imposed variance structure proportional to an empirical relatedness matrix. Lastly, conceptual and analytical similarities to and distinctions from other approaches are emphasized throughout.

RESULTS

Our simulations show that the power performance of the proposed test is quite promising compared to the considered competing strategies. The power gains are especially large when small differential differences between cases and controls are present; a likely scenario when public controls are used in multiple studies.

CONCLUSION

The proposed modified approach attains high power more consistently than that of the existing commonly implemented tests. Its performance improvement is most apparent when small but detectable systematic differences between cases and controls exist.

Correcting for relatedness in Bayesian models for genomic data association analysis

For small pedigrees, the issue of correcting for known or estimated relatedness structure in population-based Bayesian multilocus association analysis is considered. Two such relatedness corrections: [1] a random term arising from the infinite polygenic model and [2] a fixed covariate following the class D model of Bonney, are compared with the case of no correction using both simulated and real marker and gene-expression data from lymphoblastoid cell lines from four CEPH families. This comparison is performed with clinical quantitative trait locus (cQTL) models-multilocus association models where marker data and expression levels of gene transcripts as well as possible genotype x expression interaction terms are jointly used to explain quantitative trait variation. We found out that regardless of having a correction term in the model, the cQTL-models fit a few extra small-effect components (similar to finite polygenic models) which itself serves as a relatedness correction. For small data and small heritability one may use the covariate model, which clearly outperforms the infinite polygenic model in small data examples.

Common variants in the SLCO1B3 locus are associated with bilirubin levels and unconjugated hyperbilirubinemia

Bilirubin, resulting largely from the turnover of hemoglobin, is found in the plasma in two main forms: unconjugated or conjugated with glucuronic acid. Unconjugated bilirubin is transported into hepatocytes. There, it is glucuronidated by UGT1A1 and secreted into the bile canaliculi. We report a genome wide association scan in 4300 Sardinian individuals for total serum bilirubin levels. In addition to the two known loci previously involved in the regulation of bilirubin levels, UGT1A1 (P = 6.2 x 10(-62)) and G6PD (P = 2.5 x 10(-8)), we observed a strong association on chromosome 12 within the SLCO1B3 gene (P = 3.9 x 10(-9)). Our findings were replicated in an independent sample of 1860 Sardinians and in 832 subjects from the Old Order Amish (combined P < 5 x 10(-14)). We also show that SLC01B3 variants contribute to idiopathic mild unconjugated hyperbilirubinemia. Thus, SLC01B3 appears to be involved in the regulation of serum bilirubin levels in healthy individuals and in some bilirubin-related disorders that are only partially explained by other known gene variants.