The power to detect linkage disequilibrium with quantitative traits in selected samples

Genetic Influences on Cognitive Dysfunction in Schizophrenia

Schizophrenia is a severe and debilitating psychotic disorder that is highly heritable and relatively common in the population. The clinical heterogeneity associated with schizophrenia is substantial, with patients exhibiting a broad range of deficits and symptom severity. Large-scale genomic studies employing a case-control design have begun to provide some biological insight. However, this strategy combines individuals with clinically diverse symptoms and ignores the genetic risk that is carried by many clinically unaffected individuals. Consequently, the majority of the genetic architecture underlying schizophrenia remains unexplained, and the pathways by which the implicated variants contribute to the clinically observable signs and symptoms are still largely unknown. Parsing the complex, clinical phenotype of schizophrenia into biologically relevant components may have utility in research aimed at understanding the genetic basis of liability. Cognitive dysfunction is a hallmark symptom of schizophrenia that is associated with impaired quality of life and poor functional outcome. Here, we examine the value of quantitative measures of cognitive dysfunction to objectively target the underlying neurobiological pathways and identify genetic variants and gene networks contributing to schizophrenia risk. For a complex disorder, quantitative measures are also more efficient than diagnosis, allowing for the identification of associated genetic variants with fewer subjects. Such a strategy supplements traditional analyses of schizophrenia diagnosis, providing the necessary biological insight to help translate genetic findings into actionable treatment targets. Understanding the genetic basis of cognitive dysfunction in schizophrenia may thus facilitate the development of novel pharmacological and procognitive interventions to improve real-world functioning.

Between-tumor and within-tumor heterogeneity in invasive potential

For women with access to healthcare and early detection, breast cancer deaths are caused primarily by metastasis rather than growth of the primary tumor. Metastasis has been difficult to study because it happens deep in the body, occurs over years, and involves a small fraction of cells from the primary tumor. Furthermore, within-tumor heterogeneity relevant to metastasis can also lead to therapy failures and is obscured by studies of bulk tissue. Here we exploit heterogeneity to identify molecular mechanisms of metastasis. We use “organoids”, groups of hundreds of tumor cells taken from a patient and grown in the lab, to probe tumor heterogeneity, with potentially thousands of organoids generated from a single tumor. We show that organoids have the character of biological replicates: within-tumor and between-tumor variation are of similar magnitude. We develop new methods based on population genetics and variance components models to build between-tumor and within-tumor statistical tests, using organoids analogously to large sibships and vastly amplifying the test power. We show great efficiency for tests based on the organoids with the most extreme phenotypes and potential cost savings from pooled tests of the extreme tails, with organoids generated from hundreds of tumors having power predicted to be similar to bulk tests of hundreds of thousands of tumors. We apply these methods to an association test for molecular correlates of invasion, using a novel quantitative invasion phenotype calculated as the spectral power of the organoid boundary. These new approaches combine to show a strong association between invasion and protein expression of Keratin 14, a known biomarker for poor prognosis, with p = 2 × 10⁻⁴⁵ for within-tumor tests of individual organoids and p < 10⁻⁶ for pooled tests of extreme tails. Future studies using these methods could lead to discoveries of new classes of cancer targets and development of corresponding therapeutics. All data and methods are available under an open source license at https://github.com/baderzone/invasion_2019.

For women with access to healthcare and early detection, breast cancer deaths are caused primarily by metastasis rather than growth of the primary tumor. Metastasis has been difficult to study because it happens deep in the body, occurs over years, and involves a small fraction of cells from the primary tumor. Furthermore, individual cells within a tumor can behave very differently, leading to failures of therapies. Here we exploit heterogeneity to develop new methods to identify molecular mechanisms of metastasis. We use “organoids”, groups of hundreds tumor cells taken from a patient and grown in the lab. Thousands of organoids can be generated from a single tumor sample to probe different regions and amplify the amount of information provided. Organoids provide information about metastasis because they vary in their ability to invade the growth medium. We introduce a new phenotype for invasion obtained by converting the boundary of an organoid into a frequency spectrum, then summing the power across all frequencies. We analyze this metastasis-related phenotype by adapting methods from population genetics that compare the most extreme siblings in a family. We analogously compare the most invasive vs. least invasive organoids from each tumor. Power calculations suggest that studies of 50–100 individuals, with 100-1000 organoids generated from each, could reveal DNA mutations and aberrant gene expression associated with invasion. We validate this approach by demonstrating strong statistical significance between invasion and protein expression of Keratin 14, a known biomarker for poor prognosis. Future studies using these methods could lead to discoveries of new classes of cancer targets and development of corresponding therapeutics.

Endophenotypes in Schizophrenia: Digging Deeper to Identify Genetic Mechanisms

Schizophrenia (SZ) is a severe psychotic disorder that is highly heritable and common in the general population. The genetic heterogeneity of SZ is substantial, with contributions from common, rare, and de novo variants, in addition to environmental factors. Large genome-wide association studies have detected many variants that are associated with SZ, yet the pathways by which these variants influence risk remain largely unknown. SZ is also clinically heterogeneous, with patients exhibiting a broad range of deficits and symptom severity that vary over the course of illness and treatment, which has complicated efforts to identify risk variants. However, the underlying brain dysfunction forms a more stable trait marker that quantitative neurocognitive and neurophysiological endophenotypes may be able to objectively measure. These endophenotypes are less likely to be heterogeneous than the disorder and provide a neurobiological context to detect risk variants and underlying pathways among genes associated with SZ diagnosis. Furthermore, many endophenotypes are translational into animal model systems, allowing for direct evaluation of the neural circuit dysfunctions and neurobiological substrates. We review a selection of the most promising SZ endophenotypes, including prepulse inhibition, mismatch negativity, oculomotor antisaccade, letter-number sequencing, and continuous performance tests. We also highlight recent findings from large consortia that suggest the potential role of genes, particularly in the neuregulin and glutamate pathways, in several of these endophenotypes. Although endophenotypes require additional time and effort to assess, the insight into the underlying neurobiology that they provide may ultimately reveal the underlying genetic architecture for SZ and suggest novel treatment targets.

Identifying Multi-Omics Causers and Causal Pathways for Complex Traits

The central dogma of molecular biology delineates a unidirectional causal flow, i.e., DNA → RNA → protein → trait. Genome-wide association studies, next-generation sequencing association studies, and their meta-analyses have successfully identified ~12,000 susceptibility genetic variants that are associated with a broad array of human physiological traits. However, such conventional association studies ignore the mediate causers (i.e., RNA, protein) and the unidirectional causal pathway. Such studies may not be ideally powerful; and the genetic variants identified may not necessarily be genuine causal variants. In this article, we model the central dogma by a mediate causal model and analytically prove that the more remote an omics level is from a physiological trait, the smaller the magnitude of their correlation is. Under both random and extreme sampling schemes, we numerically demonstrate that the proteome-trait correlation test is more powerful than the transcriptome-trait correlation test, which in turn is more powerful than the genotype-trait association test. In conclusion, integrating RNA and protein expressions with DNA data and causal inference are necessary to gain a full understanding of how genetic causal variants contribute to phenotype variations.

An unbiased index to quantify participant's phenotypic contribution to an open-access cohort

The Personal Genome Project (PGP) is an effort to enroll many participants to create an open-access repository of genome, health and trait data for research. However, PGP participants are not enrolled for studying any specific traits and participants choose the phenotypes to disclose. To measure the extent and willingness and to encourage and guide participants to contribute phenotypes, we developed an algorithm to score and rank the phenotypes and participants of the PGP. The scoring algorithm calculates the participation index (P-index) for every participant, where 0 indicates no reported phenotypes and 100 indicate complete phenotype reporting. We calculated the P-index for all 5,015 participants in the PGP and they ranged from 0 to 96.7. We found that participants mainly have either high scores (P-index > 90, 29.5%) or low scores (P-index < 10, 57.8%). While, there are significantly more males than female participants (1,793 versus 1,271), females tend to have on average higher P-indexes (P = 0.015). We also reported the P-indexes of participants based on demographics and states like Missouri and Massachusetts have better P-indexes than states like Utah and Minnesota. The P-index can therefore be used as an unbiased way to measure and rank participant's phenotypic contribution towards the PGP.

Statistical equivalent of the classical TDT for quantitative traits and multivariate phenotypes

Clinical end-point traits are usually governed by quantitative precursors. Hence, there is active research interest in developing statistical methods for association mapping of quantitative traits. Unlike population-based tests for association, family-based tests for transmission disequilibrium are protected against population stratification. In this study, we propose a logistic regression model to test the association for quantitative traits based on a trio design. We show that the method can be viewed as a direct extension of the classical transmission diequilibrium test for binary traits to quantitative traits. We evaluate the performance of our method usingextensive simulations and compare it with an existing method, family-based association test. We found that the two methods yield comparable powers if all families are considered. However, unlike FBAT, which yields an inflated rate of false positives when noninformative trios with all three individuals' heterozygous are removed, our method maintains the correct size without compromising too much on power. We show that our method can be easily modified to incorporate multivariate phenotypes. Here, we applied this method to analyse a quantitative endophenotype associated with alcoholism.

Strategies to design and analyze targeted sequencing data: cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium Targeted Sequencing Study

BACKGROUND

Genome-wide association studies have identified thousands of genetic variants that influence a variety of diseases and health-related quantitative traits. However, the causal variants underlying the majority of genetic associations remain unknown. Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium Targeted Sequencing Study aims to follow up genome-wide association study signals and identify novel associations of the allelic spectrum of identified variants with cardiovascular-related traits.

METHODS AND RESULTS

The study included 4231 participants from 3 CHARGE cohorts: the Atherosclerosis Risk in Communities Study, the Cardiovascular Health Study, and the Framingham Heart Study. We used a case-cohort design in which we selected both a random sample of participants and participants with extreme phenotypes for each of 14 traits. We sequenced and analyzed 77 genomic loci, which had previously been associated with ≥1 of 14 phenotypes. A total of 52 736 variants were characterized by sequencing and passed our stringent quality control criteria. For common variants (minor allele frequency ≥1%), we performed unweighted regression analyses to obtain P values for associations and weighted regression analyses to obtain effect estimates that accounted for the sampling design. For rare variants, we applied 2 approaches: collapsed aggregate statistics and joint analysis of variants using the sequence kernel association test.

CONCLUSIONS

We sequenced 77 genomic loci in participants from 3 cohorts. We established a set of filters to identify high-quality variants and implemented statistical and bioinformatics strategies to analyze the sequence data and identify potentially functional variants within genome-wide association study loci.

Genetic variations in interleukin-8 and interleukin-10 are associated with pain, depressed mood, and fatigue in lung cancer patients

CONTEXT

A report by the National Cancer Institute identified that an important gap in symptom research is the investigation of multiple symptoms of cancer that might identify common biological mechanisms among cancer-related symptoms.

OBJECTIVES

We applied novel statistical methods to assess whether variants of 37 inflammation genes may serve as biologic markers of risk for severe pain, depressed mood, and fatigue in non-Hispanic white patients with non-small cell lung cancer.

METHODS

Pain, fatigue, and depressed mood were assessed before cancer treatment. We used a generalized, multivariate, classification tree approach to explore the influence of single-nucleotide polymorphisms in the inflammation genes in pain, depressed mood, and fatigue in lung cancer patients.

RESULTS

Among patients with advanced-stage disease, interleukin (IL)-8-T251A was the most relevant genetic factor for pain (odds ratio [OR] = 2.18, 95% CI = 1.34-3.55, P = 0.001), depressed mood (OR = 0.37, 95% CI = 0.14-1.0), and fatigue (OR = 2.07, 95% CI = 1.16-3.70). Among those with early-stage non-small cell lung cancer, variants in the IL-10 receptor were relevant for fatigue among women. Specifically, women with Lys_Glu or Glu_Glu genotype in the IL-10 gene had a 0.49 times lower risk of severe fatigue compared with those with Lys_Lys genotype (OR = 0.49, 95% CI = 0.25-0.92, P = 0.027). Among men with early-stage lung cancer, a marginal significance was observed for IL-1A C-889T, C/T, or T/T genotypes. These men had a lower risk of severe fatigue compared with those with C/C genotype (OR = 0.38, 95% CI = 0.13-1.06).

CONCLUSION

The interaction of multiple inflammation genes, along with nongenetic factors, underlies the occurrence of symptoms. IL-8 and IL-10 may serve as potential targets for treating multiple symptoms of cancer.

Efficient simulation of epistatic interactions in case-parent trios

Statistical approaches to evaluate interactions between single nucleotide polymorphisms (SNPs) and SNP-environment interactions are of great importance in genetic association studies, as susceptibility to complex disease might be related to the interaction of multiple SNPs and/or environmental factors. With these methods under active development, algorithms to simulate genomic data sets are needed to ensure proper type I error control of newly proposed methods and to compare power with existing methods. In this paper we propose an efficient method for a haplotype-based simulation of case-parent trios when the disease risk is thought to depend on possibly higher-order epistatic interactions or gene-environment interactions with binary exposures.

Designing a GWAS: power, sample size, and data structure

In this chapter we describe a novel Bayesian approach to designing GWAS studies with the goal of ensuring robust detection of effects of genomic loci associated with trait variation.The goal of GWAS is to detect loci associated with variation in traits of interest. Finding which of 500,000-1,000,000 loci has a practically significant effect is a difficult statistical problem, like finding a needle in a haystack. We address this problem by designing experiments to detect effects with a given Bayes factor, where the Bayes factor is chosen sufficiently large to overcome the low prior odds for genomic associations. Methods are given for various possible data structures including random population samples, case-control designs, transmission disequilibrium tests, sib-based transmission disequilibrium tests, and other family-based designs including designs for plants with clonal replication. We also consider the problem of eliciting prior information from experts, which is necessary to quantify prior odds for loci. We advocate a "subjective" Bayesian approach, where the prior distribution is considered as a mathematical representation of our prior knowledge, while also giving generic formulae that allow conservative computations based on low prior information, e.g., equivalent to the information in a single sample point. Examples using R and the R packages ldDesign are given throughout.

Impact on modes of inheritance and relative risks of using extreme sampling when designing genetic association studies

Using extreme phenotypes for association studies can improve statistical power . We study the impact of using samples with extremely high or low traits on the alternative model space, the genotype relative risks, and the genetic models in association studies. We prove the following results: when the risk allele causes high-trait values, the more extreme the high traits, the larger the genotype relative risks, which is not always true for using extreme low traits; we also prove that a genetic model theoretically changes with more extreme trait except for the recessive or dominant models. Practically, however, the impact of deviations from the true genetic model at a functional locus due to selective sampling is virtually negligible. The implications of our findings are discussed. Numerical values are reported for illustrations.

Association of CDX1 binding site of periostin gene with bone mineral density and vertebral fracture risk

SUMMARY

Periostin (POSTN) as a regulator of osteoblast differentiation and bone formation may affect susceptibility to osteoporosis. This study suggests POSTN as a candidate gene for bone mineral density (BMD) variation and vertebral fracture risk, which could better our understanding about the genetic pathogenesis of osteoporosis and will be useful in clinic in the future.

INTRODUCTION

The genetic determination of osteoporosis is complex and ill-defined. Periostin (POSTN), an extracellular matrix secreted by osteoblasts and a regulator of osteoblast differentiation and bone formation, may affect susceptibility to osteoporosis.

METHODS

We adopted a tag-single nucleotide polymorphism (SNP) based association method followed by imputation-based verification and identification of a causal variant. The association was investigated in 1,572 subjects with extreme-BMD and replicated in an independent population of 2,509 subjects. BMD was measured by dual X-ray absorptiometry. Vertebral fractures were identified by assessing vertebral height from X-rays of the thoracolumbar spine. Association analyses were performed with PLINK toolset and imputation analyses with MACH software. The top imputation finding was subsequently validated by genotyping. Interactions between POSTN and another BMD-related candidate gene sclerostin (SOST) were analyzed using MDR program and validated by logistical regression analyses. The putative transcription factor binding with target sequence was confirmed by electrophoretic mobility shift assay (EMSA).

RESULTS

Several SNPs of POSTN were associated with BMD or vertebral fractures. The most significant polymorphism was rs9547970, located at the -2,327 bp upstream (P = 6.8 × 10(-4)) of POSTN. Carriers of the minor allele G per copy of rs9547970 had 1.33 higher risk of vertebral fracture (P = 0.007). An interactive effect between POSTN and SOST upon BMD variation was suggested (P < 0.01). A specific binding of CDX1 to the sequence of POSTN with the major allele A of rs9547970 but not the variant G allele was confirmed by EMSA.

CONCLUSIONS

Our results suggest POSTN as a candidate gene for BMD variation and vertebral fracture risk.

A unified framework for detecting rare variant quantitative trait associations in pedigree and unrelated individuals via sequence data

OBJECTIVES

There is great interest to sequence unrelated or pedigree samples for detecting rare variant quantitative trait associations. In order to reduce the cost of sequencing and improve power, many studies sequence selected samples with extreme traits. Existing methods for detecting rare variant associations were developed for unrelated samples. Methods are needed to analyze (selected or randomly ascertained) pedigree samples.

METHODS

We propose a unified framework of modeling extreme trait genetic associations (MEGA) with rare variants. Using MEGA and appropriate permutation algorithms, many rare variant tests can be extended to family data. As an application, we compared study designs using both sib-pairs and unrelated individuals. Extensive simulations were carried out using realistic population genetic and complex trait models.

RESULTS

It is demonstrated that when extreme sampling is implemented within equal-sized cohorts of unrelated individuals or sib-pairs, analyzing unrelated individuals is consistently more powerful than studying sib-pairs. A higher portion of rare variants can be identified through sequencing unrelated samples compared to sibs. Alternatively, if samples are ascertained using fixed thresholds from an infinite-sized population, sequencing one sib with the most extreme trait from each extreme concordant sib-pair is consistently the most powerful design.

CONCLUSIONS

MEGA will play an important role in the analysis of sequence-based genetic association studies.

Common variants show predicted polygenic effects on height in the tails of the distribution, except in extremely short individuals

Common genetic variants have been shown to explain a fraction of the inherited variation for many common diseases and quantitative traits, including height, a classic polygenic trait. The extent to which common variation determines the phenotype of highly heritable traits such as height is uncertain, as is the extent to which common variation is relevant to individuals with more extreme phenotypes. To address these questions, we studied 1,214 individuals from the top and bottom extremes of the height distribution (tallest and shortest ∼1.5%), drawn from ∼78,000 individuals from the HUNT and FINRISK cohorts. We found that common variants still influence height at the extremes of the distribution: common variants (49/141) were nominally associated with height in the expected direction more often than is expected by chance (p<5×10⁻²⁸), and the odds ratios in the extreme samples were consistent with the effects estimated previously in population-based data. To examine more closely whether the common variants have the expected effects, we calculated a weighted allele score (WAS), which is a weighted prediction of height for each individual based on the previously estimated effect sizes of the common variants in the overall population. The average WAS is consistent with expectation in the tall individuals, but was not as extreme as expected in the shortest individuals (p<0.006), indicating that some of the short stature is explained by factors other than common genetic variation. The discrepancy was more pronounced (p<10⁻⁶) in the most extreme individuals (height<0.25 percentile). The results at the extreme short tails are consistent with a large number of models incorporating either rare genetic non-additive or rare non-genetic factors that decrease height. We conclude that common genetic variants are associated with height at the extremes as well as across the population, but that additional factors become more prominent at the shorter extreme.

Although there are many loci in the human genome that have been discovered to be significantly associated with height, it is unclear if these loci have similar effects in extremely tall and short individuals. Here, we examine hundreds of extremely tall and short individuals in two population-based cohorts to see if these known height determining loci are as predictive as expected in these individuals. We found that these loci are generally as predictive of height as expected in these individuals but that they begin to be less predictive in the most extremely short individuals. We showed that this result is consistent with models that not only include the common variants but also multiple low frequency genetic variants that substantially decrease height. However, this result is also consistent with non-additive genetic effects or rare non-genetic factors that substantially decrease height. This finding suggests the possibility of a major role of low frequency variants, particularly in individuals with extreme phenotypes, and has implications on whole-genome or whole-exome sequencing efforts to discover rare genetic variation associated with complex traits.

Identification of genes for bone mineral density variation by computational disease gene identification strategy

We previously used five freely available bioinformatics tools (Prioritizer, Geneseeker, PROSPECTR and SUSPECTS, Disease Gene Prediction, and Endeavour) to analyze the thirteen well-replicated osteoporosis susceptibility loci and identify a subset of most likely candidate osteoporosis susceptibility genes (Huang et al. in J Hum Genet 53:644-655, 2008). In the current study, we experimentally tested the association between bone mineral density (BMD) and the 9 most likely candidate genes [LAMC2(1q25-q31), MATN3(2p24-p23), ITGAV(2q31-q32), ACVR1(2q23-q24), TDGF1(3p21.31), EGF(4q25), IGF1(12q22-q23), ZIC2(13q32), BMP2(20p12)] which were pinpointed by 4 or more bioinformatics tools. Forty tag SNPs in nine candidate genes were genotyped in a southern Chinese female case-control cohort consisting of 1643 subjects. Single- and multi-marker association analyses were performed using logistic regression analysis implemented by PLINK. Potential transcription factor binding sites were predicted by MatInspector. The strongest association was observed between rs10178256 (MATN3) and trochanter (P < 0.001) and total hip BMD (P = 0.002). The SNP rs6214 (IGF1) showed consistent association with BMD at all the four measured skeletal sites (P = 0.005-0.044). Prediction of transcription factor binding suggested that the minor allele G of rs10178256 might abolish the binding of MESP1 and MESP2 which play vital roles in bone homeostasis, whereas the minor allele G of rs6214 might create an additional binding site for XBP1, a constitutive regulator of endoplasmic reticulum stress response. Our data suggested that variants in MATN3 and IGF1 were involved in BMD regulation in southern Chinese women.

Genome-wide population-based association study of extremely overweight young adults--the GOYA study

Background

Thirty-two common variants associated with body mass index (BMI) have been identified in genome-wide association studies, explaining ∼1.45% of BMI variation in general population cohorts. We performed a genome-wide association study in a sample of young adults enriched for extremely overweight individuals. We aimed to identify new loci associated with BMI and to ascertain whether using an extreme sampling design would identify the variants known to be associated with BMI in general populations.

Methodology/Principal Findings

From two large Danish cohorts we selected all extremely overweight young men and women (n = 2,633), and equal numbers of population-based controls (n = 2,740, drawn randomly from the same populations as the extremes, representing ∼212,000 individuals). We followed up novel (at the time of the study) association signals (p<0.001) from the discovery cohort in a genome-wide study of 5,846 Europeans, before attempting to replicate the most strongly associated 28 SNPs in an independent sample of Danish individuals (n = 20,917) and a population-based cohort of 15-year-old British adolescents (n = 2,418). Our discovery analysis identified SNPs at three loci known to be associated with BMI with genome-wide confidence (P<5×10⁻⁸; FTO, MC4R and FAIM2). We also found strong evidence of association at the known TMEM18, GNPDA2, SEC16B, TFAP2B, SH2B1 and KCTD15 loci (p<0.001), and nominal association (p<0.05) at a further 8 loci known to be associated with BMI. However, meta-analyses of our discovery and replication cohorts identified no novel associations.

Significance

Our results indicate that the detectable genetic variation associated with extreme overweight is very similar to that previously found for general BMI. This suggests that population-based study designs with enriched sampling of individuals with the extreme phenotype may be an efficient method for identifying common variants that influence quantitative traits and a valid alternative to genotyping all individuals in large population-based studies, which may require tens of thousands of subjects to achieve similar power.

Combined analysis of three genome-wide association studies on vWF and FVIII plasma levels

Background

Elevated levels of factor VIII (FVIII) and von Willebrand Factor (vWF) are well-established risk factors for cardiovascular diseases, in particular venous thrombosis. Although high, the heritability of these traits is poorly explained by the genetic factors known so far. The aim of this work was to identify novel single nucleotide polymorphisms (SNPs) that could influence the variability of these traits.

Methods

Three independent genome-wide association studies for vWF plasma levels and FVIII activity were conducted and their results were combined into a meta-analysis totalling 1,624 subjects.

Results

No single nucleotide polymorphism (SNP) reached the study-wide significance level of 1.12 × 10^-7 that corresponds to the Bonferroni correction for the number of tested SNPs. Nevertheless, the recently discovered association of STXBP5, STX2, TC2N and CLEC4M genes with vWF levels and that of SCARA5 and STAB2 genes with FVIII levels were confirmed in this meta-analysis. Besides, among the fifteen novel SNPs showing promising association at p < 10^-5 with either vWF or FVIII levels in the meta-analysis, one located in ACCN1 gene also showed weak association (P = 0.0056) with venous thrombosis in a sample of 1,946 cases and 1,228 controls.

Conclusions

This study has generated new knowledge on genomic regions deserving further investigations in the search for genetic factors influencing vWF and FVIII plasma levels, some potentially implicated in VT, as well as providing some supporting evidence of previously identified genes.

A simple bias correction in linear regression for quantitative trait association under two-tail extreme selection

Selective genotyping can increase power in quantitative trait association. One example of selective genotyping is two-tail extreme selection, but simple linear regression analysis gives a biased genetic effect estimate. Here, we present a simple correction for the bias.

Bivariate association analysis in selected samples: application to a GWAS of two bone mineral density phenotypes in males with high or low BMD

Our specific aims were to evaluate the power of bivariate analysis and to compare its performance with traditional univariate analysis in samples of unrelated subjects under varying sampling selection designs. Bivariate association analysis was based on the seemingly unrelated regression (SUR) model that allows different genetic models for different traits. We conducted extensive simulations for the case of two correlated quantitative phenotypes, with the quantitative trait locus making equal or unequal contributions to each phenotype. Our simulation results confirmed that the power of bivariate analysis is affected by the size, direction and source of the phenotypic correlations between traits. They also showed that the optimal sampling scheme depends on the size and direction of the induced genetic correlation. In addition, we demonstrated the efficacy of SUR-based bivariate test by applying it to a real Genome-Wide Association Study (GWAS) of Bone Mineral Density (BMD) values measured at the lumbar spine (LS) and at the femoral neck (FN) in a sample of unrelated males with low BMD (LS Z-scores ≤ -2) and with high BMD (LS and FN Z-scores >0.5). A substantial amount of top hits in bivariate analysis did not reach nominal significance in any of the two single-trait analyses. Altogether, our studies suggest that bivariate analysis is of practical significance for GWAS of correlated phenotypes.

A flexible model for association analysis in sibships with missing genotype data

A common design in family-based association studies consists of siblings without parents. Several methods have been proposed for analysis of sibship data, but they mostly do not allow for missing data, such as haplotype phase or untyped markers. On the other hand, general methods for nuclear families with missing data are computationally intensive when applied to sibships, since every family has missing parents that could have many possible genotypes. We propose a computationally efficient model for sibships by conditioning on the sets of alleles transmitted into the sibship by each parent. This means that the likelihood can be written only in terms of transmitted alleles and we do not have to sum over all possible untransmitted alleles when they cannot be deduced from the siblings. The model naturally accommodates missing data and admits standard theory of estimation, testing, and inclusion of covariates. Our model is quite robust to population stratification and can test for association in the presence of linkage. We show that our model has similar power to FBAT for single marker analysis and improved power for haplotype analysis. Compared to summing over all possible untransmitted alleles, we achieve similar power with considerable reductions in computation time.

Coronary artery disease: an example case study

This chapter illustrates various general issues in genetic epidemiology in relation to coronary artery disease (CAD). This is a disease strongly influenced by environmental/lifestyle factors, such as smoking, but with substantial estimated heritability. Researchers aiming to identify susceptibility genes have used several different definitions of CAD, some focusing on the common presentation of myocardial infarction (MI) and others adopting broader criteria, often imposing an upper limit to age at diagnosis to minimise environmental effects. Many candidate gene association studies and a few large genome-wide linkage studies have been conducted, but with limited success. Several heritable quantitative traits are strongly related to risk of CAD (e.g. blood pressure and cholesterol levels), and much research has been focussed on identifying genes that influence these traits. Quantitative traits have the advantage of being measurable on any individual, allowing them to be studied in population-based cohorts. However, they also tend to vary considerably over time, and intra-individual variation needs to be taken into account in analyses. In the last few years, both CAD itself and related quantitative traits have been studied in genome-wide association studies using large sample sizes. Several novel genetic loci influencing CAD have been identified and replicated, in addition to many loci influencing related quantitative traits. However, despite this recent success, only a small fraction of the genetic contribution to risk has been explained.

Common variants in FLNB/CRTAP, not ARHGEF3 at 3p, are associated with osteoporosis in southern Chinese women

SUMMARY

We performed an association study of five candidate genes within chromosome 3p14-25 in 1,080 Chinese female subjects. Polymorphisms in FLNB/CRTAP are associated with bone mineral density (BMD) in Chinese.

INTRODUCTION

Chromosomal region 3p14-25 has shown strong evidence of linkage to BMD in genome-wide linkage scans. The variants responsible for this linkage signal, nonetheless, remain obscure.

METHODS

Thirty SNPs in five positional and functional candidate genes within 3p14-25 (PPARG, CRTAP, TDGF1, PTHR1, and FLNB) and rs7646054 in the ARHGEF3 gene were genotyped in a case-control cohort of 1,080 Chinese females. Allelic and haplotypic association were tested using logistic regression analysis implemented in PLINK software. Potential transcription factor binding sites were predicted with MatInspector.

RESULTS

Multiple SNPs and haplotypes in FLNB were significantly associated with BMDs, with the strongest association between lumbar spine BMD and rs9828717 (p = 0.005). SNP rs7623768 and the haplotype G-C of rs4076086-rs7623768 in CRTAP were associated with femoral neck BMD (p = 0.009 and p = 0.003, respectively). PTHR1 showed haplotypic associations with lumbar spine and femoral neck BMD (p = 0.02 and p = 0.044, respectively). Nevertheless, the association between rs7646054 in ARHGEF3 and BMD observed in Caucasians was not replicated in our samples. Comparative genomics analysis indicated that rs9828717 is located within a highly conserved region. The minor T allele at rs9828717 may lead to loss of binding site for nuclear factor of activated T cells which binds and triggers the transcriptional program of osteoblasts.

CONCLUSIONS

Our data suggest that variants in FLNB and CRTAP at 3p are involved in BMD regulation in southern Chinese.

The -9247 T/C polymorphism in the SOST upstream regulatory region that potentially affects C/EBPalpha and FOXA1 binding is associated with osteoporosis

Accumulating evidence shows that genes that cause monogenic diseases also contribute to similar complex disease in the general population. We sought to determine whether the allelic variation in seven monogenic bone disease genes (CLCN7, TCIRGI, SOST, CA2, CSTK, TGFB1 and SLC26A2) contributes to osteoporosis/bone mineral density (BMD) variation in the normal Chinese population. We conducted a gene-wide tag SNP-based association study in 1243 Chinese subjects with low BMD (Z-scores < or = -1.28, equivalent to the lowest 10% of the population) and high BMD (Z-score > or = +1.0). Twenty-two tag SNPs were selected and genotyped by using the high-throughput Sequenom genotyping platform. Allelic and haplotype association tests were conducted by Haploview and binary logistic regression analyses. The -9247 polymorphism rs1230399 in the upstream regulatory region of the sclerostin gene showed significant genotypic/allelic associations with spine, femoral neck, trochanter and total hip BMD (P=0.03-0.004). The T-allele of rs1230399 increased the risk of osteoporosis (OR=1.52, P=0.005). Computational analysis showed that rs1230399 is located at the core consensus recognition site of two cooperating transcription factors C/EBPalpha and FOXA1 that modulate estrogen receptor function. T-->C polymorphism abolishes the binding of both C/EBPalpha and FOXA1 to the sclerostin gene. Our data suggest a mechanistic link between rs1230399 and BMD through estrogen ERalpha/FOXA1 signaling pathways driven by long-distance enhancers.

Novel sib pair selection strategy increases power in quantitative association analysis

Quantitative-trait association studies have been widely used in search for genetic loci for complex traits in recent years. Yet, fiscal constraints still prohibit many on-going research projects from recruiting a large number of individuals for genotyping to reach a desired level of statistical power. Accordingly, in this article, we describe a novel sib pair sampling strategy for genotyping in QTL association studies. With the use of phenotypic scores (and IBD allele-sharing probabilities if available), the genetic effect of a biallelic additive trait locus can be properly modelled within the maximum-likelihood variance components framework proposed by Fulker et al. (Am J Hum Genet 64(1):259-267, 1999) and sib pairs can be rank-ordered by use of informativeness indices. The performance of our method was investigated using simulation. The power of our approach was shown to be higher when compared with other phenotypic selection schemes. An R-script implementing all the selection approaches (including the traditional phenotype-based ones) used in the simulation is available at http://statgen.hku.hk/jshkwan .

Obesity is Associated with Genetic Variants That Alter Dopamine Availability

Human and animal studies have implicated dopamine in appetite regulation, and family studies have shown that BMI has a strong genetic component. Dopamine availability is controlled largely by three enzymes: COMT, MAOA and MAOB, and by the dopamine transporter SLC6A3, and each gene has a well-characterized functional variant. Here we look at these four functional polymorphisms together, to investigate how heritable variation in dopamine levels influences the risk of obesity in a cohort of 1150, including 240 defined as obese (BMI > or = 30). The COMT and SLC6A3 polymorphisms showed no association with either weight, BMI or obesity risk. We found, however, that both MAOA and MAOB show an excess of the low-activity genotypes in obese individuals (MAOA:chi2= 15.45, p = 0.004; MAOB:chi2= 8.05, p = 0.018). Additionally, the MAOA genotype was significantly associated with both weight (p = 0.0005) and BMI (p = 0.001). When considered together, the 'at risk genotype'--low activity genotypes at both the MAOA and MAOB loci--shows a relative risk for obesity of 5.01. These results have not been replicated and, given the experience of complex trait genetics, warrant caution in interpretation. In implicating both the MAOA and MOAB variants, however, this study provides the first indication that dopamine availability (as opposed to other effects of MAOA) is involved in human obesity. It is therefore a priority to assess the associations in replication datasets.

Genetic analyses in a sample of individuals with high or low BMD shows association with multiple Wnt pathway genes

Using a moderate-sized cohort selected with extreme BMD (n = 344; absolute value BMD, 1.5-4.0), significant association of several members of the Wnt signaling pathway with bone densitometry measures was shown. This confirms that extreme truncate selection is a powerful design for quantitative trait association studies of bone phenotypes.

INTRODUCTION

Although the high heritability of BMD variation has long been established, few genes have been conclusively shown to affect the variation of BMD in the general population. Extreme truncate selection has been proposed as a more powerful alternative to unselected cohort designs in quantitative trait association studies. We sought to test these theoretical predictions in studies of the bone densitometry measures BMD, BMC, and femoral neck area, by investigating their association with members of the Wnt pathway, some of which have previously been shown to be associated with BMD in much larger cohorts, in a moderate-sized extreme truncate selected cohort (absolute value BMD Z-scores = 1.5-4.0; n = 344).

MATERIALS AND METHODS

Ninety-six tag-single nucleotide polymorphism (SNPs) lying in 13 Wnt signaling pathway genes were selected to tag common genetic variation (minor allele frequency [MAF] > 5% with an r(2) > 0.8) within 5 kb of all exons of 13 Wnt signaling pathway genes. The genes studied included LRP1, LRP5, LRP6, Wnt3a, Wnt7b, Wnt10b, SFRP1, SFRP2, DKK1, DKK2, FZD7, WISP3, and SOST. Three hundred forty-four cases with either high or low BMD were genotyped by Illumina Goldengate microarray SNP genotyping methods. Association was tested either by Cochrane-Armitage test for dichotomous variables or by linear regression for quantitative traits.

RESULTS

Strong association was shown with LRP5, polymorphisms of which have previously been shown to influence total hip BMD (minimum p = 0.0006). In addition, polymorphisms of the Wnt antagonist, SFRP1, were significantly associated with BMD and BMC (minimum p = 0.00042). Previously reported associations of LRP1, LRP6, and SOST with BMD were confirmed. Two other Wnt pathway genes, Wnt3a and DKK2, also showed nominal association with BMD.

CONCLUSIONS

This study shows that polymorphisms of multiple members of the Wnt pathway are associated with BMD variation. Furthermore, this study shows in a practical trial that study designs involving extreme truncate selection and moderate sample sizes can robustly identify genes of relevant effect sizes involved in BMD variation in the general population. This has implications for the design of future genome-wide studies of quantitative bone phenotypes relevant to osteoporosis.

A review of current approaches to identifying human genes involved in myopia

The prevalence of myopia is high in many parts of the world, particularly among the Orientals such as Chinese and Japanese. Like other complex diseases such as diabetes and hypertension, myopia is likely to be caused by both genetic and environmental factors, and possibly their interactions. Owing to multiple genes with small effects, genetic heterogeneity and phenotypic complexity, the study of the genetics of myopia poses a complex challenge. This paper reviews the current approaches to the genetic analysis of complex diseases and how these can be applied to the identification of genes that predispose humans to myopia. These approaches include parametric linkage analysis, non-parametric linkage analysis like allele-sharing methods and genetic association studies. Basic concepts, advantages and disadvantages of these approaches are discussed and explained using examples from the literature on myopia. Microsatellites and single nucleotide polymorphisms are common genetic markers in the human genome and are indispensable tools for gene mapping. High throughput genotyping of millions of such markers has become feasible and efficient with recent technological advances. In turn, this makes the identification of myopia susceptibility genes a reality.

FTO gene SNPs associated with extreme obesity in cases, controls and extremely discordant sister pairs

Background

FTO is a gene located in chromosome region 16q12.2. Recently two studies have found associations of several single nucleotide polymorphisms (SNPs) in FTO with body mass index (BMI) and obesity, particularly rs1421085, rs17817449, and rs9939609.

Methods

We examined these three SNPs in 583 extremely obese women with current BMI greater than 35 kg/m² and lifetime BMI greater than 40 kg/m², and 544 controls who were currently normal weight (BMI<25 kg/m²) and had never been overweight during their lifetimes.

Results

We detected highly significant associations of obesity with alleles in all three SNPs (p < 10^-9). The strongest association was with rs1421085 (p = 3.04 × 10^-10, OR = 1.75, CI = 1.47–2.08). A subset of 99 cases had extremely discordant sisters with BMI<25 kg/m². The discordant sisters differed in allele and genotype frequencies in parallel with the overall case and control sample. The strongest association was with rs17817449 (z = 3.57, p = 3.6 × 10^-4).

Conclusion

These results suggest common variability in FTO is associated with increased obesity risk or resistance and may in part account for differences between closely related individuals.

Analysis of quantitative trait loci

Diseases with complex inheritance are characterized by multiple genetic and environmental factors that often interact to produce clinical symptoms. In addition, etiological heterogeneity (different risk factors causing similar phenotypes) obscure the inheritance pattern among affected relatives and hamper the feasibility of gene-mapping studies. For such diseases, the careful selection of quantitative phenotypes that may represent intermediary risk factors for disease development (intermediate phenotypes) is etiologically more homogeneous than the disease per se. Over the last 15 years quantitative trait locus mapping has become a popular method for understanding the genetic basis for intermediate phenotypes. This chapter provides an introduction to classical and recent strategies for mapping quantitative trait loci in humans.

Platelet responsiveness to in vitro aspirin is independent of COX-1 and COX-2 protein levels and polymorphisms

Aspirin's inhibitory effect on platelet function has been shown to be highly heterogeneous. However, due to the considerable individual variation in pharmacokinetics after aspirin intake, it has been difficult to investigate the mechanism of aspirin resistance empirically. Our objective was to examine whether platelet responsiveness to in vitro aspirin treatment could be affected by cyclooxygenase (COX)-1/2 protein levels in platelets or single-nucleotide polymorphisms (SNPs), which could possibly change specific activity of enzymes and/or aspirin susceptibility. Collagen/epinephrine closure time (CEPI-CT) of PFA-100 in blood from 178 healthy males was assessed with/without aspirin. Platelet COX-1 protein levels and the sequences of COX-1 gene exons were examined in three groups categorized by CEPI-CT: PR (Poor responders to aspirin), 10 people showing the shortest CEPI-CT under aspirin; GR-High or GR-Low (good responders to aspirin with high or low platelet basal reactivity), 10 people showing CEPI-CT over 300 s under aspirin and having the shortest or longest basal CEPI-CT, respectively. We analyzed the three groups, representing phenotypic extremes, aiming to increase statistical power to investigate the possible relevance of COXs to platelet response to aspirin. Western blot analysis revealed that COX-1 was abundantly expressed in platelets at comparable levels among the three groups, whereas COX-2 was undetectable. The frequencies of nonsynonymous COX-1/2 SNPs were unlikely to explain the difference in aspirin responsiveness considering the observed genotype frequencies and wide individual variation in platelet response. These results suggest that heterogeneity in platelet responsiveness to in vitro aspirin is independent of COX-1/2 protein levels and SNPs.

Investigating gene environment interaction in complex diseases: increasing power by selective sampling for environmental exposure

BACKGROUND

The often limited influence of disease associated alleles on the vulnerability to complex diseases has lead to increased interest in environmental interaction with genotype. However, gene environmental interactions (GEIs) are not easily studied, since high numbers of subjects are required to detect GEI.

METHODS AND RESULTS

This study provides a potential useful method to increase the power of such studies through selective sampling for environmental exposure. We show that selecting the top and bottom 10% regarding environmental exposure can lead to a 70% reduction in the required number of subjects for genotyping.

CONCLUSION

This study demonstrates the potential usefulness of selective sampling in the study of the interplay between genes and environment. The reduction of required subjects can be particularly advantageous in studies where genotyping is extensive, such as in whole genome screens or in studies where phenotyping is expensive.

The Consortium on the Genetics of Endophenotypes in Schizophrenia: model recruitment, assessment, and endophenotyping methods for a multisite collaboration

BACKGROUND

The Consortium on the Genetics of Schizophrenia (COGS) is an ongoing, National Institute of Mental Health-funded, 7-site collaboration investigating the occurrence and genetic architecture of quantitative endophenotypes related to schizophrenia. The purpose of this article is to provide a description of the COGS structure and methods, including participant recruitment and assessment.

METHODS

The hypothesis-driven recruitment strategy ascertains families that include a proband with a Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition diagnosis of schizophrenia, and at least one unaffected full sibling available for genotyping and endophenotyping, along with parents available for genotyping and (optional depending on age) endophenotyping. The family structure is selected to provide contrast in quantitative endophenotypic traits and thus to maximize the power of the planned genetic analyses. Probands are recruited from many sources including clinician referrals, local National Alliance for the Mentally Ill chapters, and advertising via the media. All participants undergo a standardized protocol that includes clinical characterization, a blood draw for genotyping, and endophenotype assessments (P50 suppression, prepulse inhibition, antisaccade performance, continuous performance tasks, letter-number span, verbal memory, and a computerized neurocognitive battery). Investigators participate in weekly teleconferences to coordinate and evaluate recruitment, clinical assessment, endophenotyping, and continuous quality control of data gathering and analyses. Data integrity is maintained through use of a highly quality-assured, centralized web-based database.

RESULTS

As of February 2006, 355 families have been enrolled and 688 participants have been endophenotyped, including schizophrenia probands (n = 154, M:F = 110:44), first-degree biological relatives (n = 343, M:F = 151:192), and community comparison subjects (n = 191, M:F = 81:110).

DISCUSSION

Successful multisite genetics collaborations must institute standardized methodological criteria for assessment and recruitment that are clearly defined, well communicated, and uniformly applied. In parallel, studies utilizing endophenotypes require strict adherence to criteria for cross-site data acquisition, equipment calibration and testing and software equivalence, and continuous quality assurance for many measures obtained across sites. This report describes methods and presents the structure of the COGS as a model of multisite endophenotype genetic studies. It also provides demographic information after the first 2 years of data collection on a sample for whom the behavioral data and genetics of endophenotype performance will be fully characterized in future articles. Some issues discussed in the reviews that follow reflect the challenges of evaluating endophenotypes in studies of the genetic architecture of endophenotypes in schizophrenia.

Deconstructing schizophrenia: an overview of the use of endophenotypes in order to understand a complex disorder

The genetics of schizophrenia has been approached utilizing a variety of methods. One emerging strategy is the use of endophenotypes in order to understand and identify the functional importance of genetically transmitted, brain-based deficits across schizophrenia kindreds. The endophenotype strategy is a topic of this issue of Schizophrenia Bulletin. Endophenotypes are quantitative, heritable, trait-related deficits typically assessed by laboratory-based methods rather than clinical observation. Endophenotypes are seen as closer to genetic variation than are clinical symptoms of schizophrenia, and are therefore closely linked to heritable risk factors. There has been a broad expansion of opportunities available to psychiatric neuroscientists who use the endophenotype strategy to understand the genetic basis of schizophrenia. In this context, genetic variation such as single nucleotide polymorphisms (SNPs) induces abnormalities in endophenotypic domains such as neurocognition, neurodevelopment, metabolism, and neurophysiology. This article discusses the challenges that abound in genetic research of schizophrenia, including issues in ascertainment, epistasis, ethnic diversity, and the potentially normalizing effects of second-generation antipsychotic medications on neurocognitive and neurophysiological measures. Robust strategies for meeting these challenges are discussed in this review and the subsequent articles in this issue. This article summarizes conceptual advances and progress in the measurement and use of endophenotypes in schizophrenia that form the basis of the multisite National Institute of Mental Health Consortium on the Genetics of Schizophrenia. The endophenotype strategy offers powerful and exciting opportunities to understand the genetically conferred neurobiological vulnerabilities and possible new strong inference and molecularly based treatments for schizophrenia.

Statistical power of association using the extreme discordant phenotype design

BACKGROUND

Selective genotyping has been proven to be an effective design for mapping quantitative trait loci (QTL), either by linkage or by allelic association, wherein the individual trait values can be used as the indices for phenotype selection. It has also been proposed that association studies of dichotomous traits can benefit from such design. When there is no quantitative measurement for phenotype available, cases and/or controls having extreme discordant phenotypes (EDP) can still be selected, based on their exposure status to a drug toxicity or environmental risk factor. The advantage of EDP design is intuitive and it has been successfully used in a number of studies.

METHODS

In this report, we developed a statistical method to calculate the power of EDP methodology, using a mixture model of genotype-specific distributions of a single biallelic susceptibility locus. We also compared the power of three statistical tests commonly used in association studies - including the chi test of allelic frequencies, the chi test of genotype frequencies, and the Armitage trend test. The power of two different EDP designs was evaluated under a range of scenarios.

RESULTS AND CONCLUSION

Our results indicate that the chi test of genotype frequency is a robust, though less powerful, test for single-locus association, and that EDP methodology is a powerful design for genetic association studies - especially those of common diseases caused by quantifiable drug toxicity or environmental risk factors.

Tests of linkage and association of PTH/PTHrP receptor type 1 gene with bone mineral density and height in Caucasians

Parathyroid hormone/parathyroid hormone-related peptide receptor type 1 (PTHR1) plays an important role in calcium metabolism. It was previously shown to influence variation in bone mineral density (BMD). To investigate its importance in a typical U.S. Caucasian population, we tested linkage or association of the PTHR1 gene with BMD and height. Altogether, 1873 subjects from 405 Caucasian nuclear families were studied. BMD was measured at the lumbar spine (L1-L4) and total hip (femoral neck, trochanter, and intertrochanter regions). Four single nucleotide polymorphisms (SNPs) in the PTHR1 gene were genotyped. Sixteen haplotypes were reconstructed. Only two major haplotypes had frequencies >3% and were thus used for the analysis. Analyses were performed for BMD and height in the total sample and for peak BMD (PBMD) achieved in offspring subjects aged 20-50 in a subsample of 387 nuclear families. We found suggestive evidence for total association between haplotype 13 (AATG) and hip PBMD (P = 0.031). For height, evidence of within-family association was suggested for SNP1, SNP2, and haplotype 4 (GGCA) (P < or = 0.05). Our findings suggest that the PTHR1 gene may be important for PBMD, height variation, or both, although the significance is dampened by correction for multiple testing.

High-resolution association mapping of quantitative trait loci: a population-based approach

In this article, population-based regression models are proposed for high-resolution linkage disequilibrium mapping of quantitative trait loci (QTL). Two regression models, the "genotype effect model" and the "additive effect model," are proposed to model the association between the markers and the trait locus. The marker can be either diallelic or multiallelic. If only one marker is used, the method is similar to a classical setting by Nielsen and Weir, and the additive effect model is equivalent to the haplotype trend regression (HTR) method by Zaykin et al. If two/multiple marker data with phase ambiguity are used in the analysis, the proposed models can be used to analyze the data directly. By analytical formulas, we show that the genotype effect model can be used to model the additive and dominance effects simultaneously; the additive effect model takes care of the additive effect only. On the basis of the two models, F-test statistics are proposed to test association between the QTL and markers. By a simulation study, we show that the two models have reasonable type I error rates for a data set of moderate sample size. The noncentrality parameter approximations of F-test statistics are derived to make power calculation and comparison. By a simulation study, it is found that the noncentrality parameter approximations of F-test statistics work very well. Using the noncentrality parameter approximations, we compare the power of the two models with that of the HTR. In addition, a simulation study is performed to make a comparison on the basis of the haplotype frequencies of 10 SNPs of angiotensin-1 converting enzyme (ACE) genes.

Categorical and continuous models of liability to externalizing disorders: a direct comparison in NESARC

CONTEXT

Patterns of genetic, environmental, and phenotypic relationships among antisocial behavior and substance use disorders indicate the presence of a common externalizing liability. However, whether this liability is relatively continuous and graded, or categorical and class-like, has not been well established.

OBJECTIVES

To compare the fit of categorical and continuous models of externalizing liability in a large, nationally representative sample.

DESIGN

Categorical and continuous models of externalizing liability were compared using interview data from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC).

SETTING

Face-to-face interviews conducted in the United States.

PARTICIPANTS

Random sample of 43 093 noninstitutionalized adult civilians living in the United States.

MAIN OUTCOME MEASURES

Lifetime and current (past 12 months) diagnoses of antisocial personality disorder, nicotine dependence, alcohol dependence, marijuana dependence, cocaine dependence, and other substance dependence.

RESULTS

In the entire sample, as well as for males and females separately, using either lifetime or current diagnoses, the best-fitting model of externalizing liability was a continuous normal model. Moreover, there was a general trend toward latent trait models fitting better than latent class models, indicating that externalizing liability was continuous and graded, rather than categorical and class-like.

CONCLUSIONS

Liability to externalizing spectrum disorders is graded and continuous normal in distribution. Research regarding etiology, assessment, and treatment of externalizing disorders should target externalizing liability over a range of severity. Current diagnoses represent extremes of this continuous liability distribution, indicating that conditions currently classified as subthreshold are likely to provide important information regarding liability to externalizing phenomena.

Complex phenotypes and complex genetics: an introduction to genetic studies of complex traits

There is currently intense interest in the genetic factors contributing to many diseases with cardiovascular complications. Diseases like atherosclerosis, diabetes, and hypertension are referred to as complex traits because multiple genes contribute to the phenotype either individually or through interactions with each other or the environment. Enabled and energized by the striking successes over the past 20 years in identifying genes that are responsible for single gene traits, many geneticists have turned to the investigation of methods that will allow for the dissection of complex traits. There have already been some successes, so there is no reason to consider the problem as inherently intractable. However, it is important to reflect on what conditions are necessary for the identification of genes that operate in complex traits. A recurring theme in this research area has been difficulty in repeating and validating research findings, and this most often can be attributed to limitations in study design. It is also important to consider that any particular research strategy can only hope to describe a portion of factors that contribute to variation in the population; therefore, the genetic approach cannot be a panacea. New efficient technologies for genotyping and public databases describing the fine structure of genetic correlations in the genome should aid many aspects of the gene discovery process.

A comparison of tagging methods and their tagging space

Single-nucleotide polymorphism (SNP) tagging is widely used as a way of saving genotyping costs in association studies. A number of different tagging methods have been developed to reduce the number of markers to be genotyped while maintaining power for detecting effects on non-assayed SNPs. How the different methods perform in different settings, the degree to which they overlap and share common tags and how they differ are important questions. We investigated these questions by comparing three widely used tagging methods/algorithms--one haplotype r2-based method, one pair-wise r2-based method and one method which was based on haplotype diversity but focused on major haplotypes. Tagging efficiency was defined as the number of genotyped markers divided by the number of tagging SNPs. Tagging effectiveness was defined as the proportion of un-genotyped or 'hidden' SNPs being detected (having a pair-wise or haplotype r2 with a set of tagging SNPs over a threshold, e.g. haplotype r2> or =0.80). The ENCODE regions genotyped on the HapMap CEPH individuals were examined in this study. Tagging effectiveness was generally poor for rare SNPs than for common SNPs, for all three tagging methods. Inclusion of rare SNPs into initial HapMap scheme could enhance the performance of tags on rare hidden SNPs at the expense of increased genotyping cost. At a moderate tagging efficiency, more than 90% of hidden SNPs detected by tagging SNPs selected by one method were also detected by tagging SNPs selected by another method, and this figure could be increased to 100% if tagging efficiency was allowed to drop. These results indicate that the tagging space is highly concordant between different tagging methods, despite the fact that they often involve different sets of tagging SNPs.

Mapping quantitative trait loci in humans: achievements and limitations

Recent advances in statistical methods and genomic technologies have ushered in a new era in mapping clinically important quantitative traits. However, many refinements and novel statistical approaches are required to enable greater successes in this mapping. The possible impact of recent findings pertaining to the structure of the human genome on efforts to map quantitative traits is yet unclear.

Pedigree linkage disequilibrium mapping of quantitative trait loci

In this paper, we propose to use pedigrees of any size and any types of relatives in joint high-resolution linkage disequilibrium (LD) and linkage mapping of quantitative trait loci (QTL) by variance component models. Two or multiple markers can be simultaneously used in modeling association with the trait locus, instead of using one marker a time in the analysis. The proposed method can provide a unified result by using two or multiple markers in the modeling. This may avoid the complications of different results obtained from the separate analysis of marker by marker. The models simultaneously incorporate both linkage and LD information. The measures of LD are modeled by mean coefficients, and linkage information is modeled by variance-covariance matrix. Using analytical formulas to calculate the regression coefficients, the genetic effects are shown to be decomposed into additive and dominance components. The noncentrality parameter approximations of test statistics of LD are provided to make power calculations. Power and type I error rates are explored to investigate the merit of the proposed method by both the analytical formulas and simulations. Comparing with the association between-family and association within-family ('AbAw') approach of Fulker and Abecasis et al, it is evident that the method proposed in this article is more powerful. The method is applied to investigate the relation between polymorphisms in the angiotensin 1-converting enzyme (ACE) genes and circulating ACE levels, with a better result than that of the 'AbAw' approach. Moreover, two markers I/D and 4656(CT)3/2 can fully interpret association with the trait locus at a 0.01 significance level, which provides a unique result for the ACE data.

Combined linkage and association mapping of quantitative trait loci by multiple markers

Using multiple diallelic markers, variance component models are proposed for high-resolution combined linkage and association mapping of quantitative trait loci (QTL) based on nuclear families. The objective is to build a model that may fully use marker information for fine association mapping of QTL in the presence of prior linkage. The measures of linkage disequilibrium and the genetic effects are incorporated in the mean coefficients and are decomposed into orthogonal additive and dominance effects. The linkage information is modeled in variance-covariance matrices. Hence, the proposed methods model both association and linkage in a unified model. On the basis of marker information, a multipoint interval mapping method is provided to estimate the proportion of allele sharing identical by descent (IBD) and the probability of sharing two alleles IBD at a putative QTL for a sib-pair. To test the association between the trait locus and the markers, both likelihood-ratio tests and F-tests can be constructed on the basis of the proposed models. In addition, analytical formulas of noncentrality parameter approximations of the F-test statistics are provided. Type I error rates of the proposed test statistics are calculated to show their robustness. After comparing with the association between-family and association within-family (AbAw) approach by Abecasis and Fulker et al., it is found that the method proposed in this article is more powerful and advantageous based on simulation study and power calculation. By power and sample size comparison, it is shown that models that use more markers may have higher power than models that use fewer markers. The multiple-marker analysis can be more advantageous and has higher power in fine mapping QTL. As an application, the Genetic Analysis Workshop 12 German asthma data are analyzed using the proposed methods.

Association between evolutionary history of angiotensinogen haplotypes and plasma levels

Over the last decade, considerable effort has been invested in studying the associations between angiotensinogen (AGT) variants, AGT plasma levels and high blood pressure. Evidence accumulated to date consistently supports the relationship between the AGT locus and the protein level, while an influence on blood pressure has been difficult to establish; in both instances the predisposing molecular variants are not fully defined. An evolutionary approach, taking into account the phylogenetic relationship between all the polymorphisms at this locus, may improve our understanding of the genetic nature of these quantitative phenotypes. Accordingly we sequenced a 6.8 kb region of the AGT gene in 57 Nigerian individuals (29 with high AGT plasma levels and 28 with low AGT plasma levels). Haplotypes were grouped into seven major haplogroups and their phylogenetic relationship was established. The association between haplogroups and AGT plasma levels was investigated. A significant linear correlation was detected between haplogroup genetic distance and AGT levels, suggesting a nonrandom accumulation of risk-associated mutations during the evolutionary history of the AGT gene.

High-resolution joint linkage disequilibrium and linkage mapping of quantitative trait loci based on sibship data

This paper proposes variance component models for high resolution joint linkage disequilibrium (LD) and linkage mapping of quantitative trait loci (QTL) based on sibship data; this can include population data if independent individuals are treated as single sibships. One application of these models is late onset complex disease gene mapping, when parental data are not available. The models simultaneously incorporate both LD and linkage information. The LD information is contained in mean coefficients of sibship data. The linkage information is contained in the variance-covariance matrices of trait values for sibships with at least two siblings. We derive formulas for calculating the probability of sharing two trait alleles identical by descent (IBD) for sibpairs in interval mapping of QTL; this is the coefficient of dominant variance of the trait covariance of sibpairs on major QTL. To investigate the performance of the formulas, we calculate the numerical values via the formulas and get satisfactory approximations. We compare the power and sample sizes for both LD and linkage mapping. By simulation and theoretical analysis, we compare the results with those of Fulker and Abecasis "AbAw" approach. It is well known that the resolution of linkage analysis can be low for complex disease gene mapping. LD mapping, on the other hand, can increase mapping precision and is useful in high resolution mapping. Linkage analysis is less sensitive to population subdivisions and admixtures. The level of LD is sensitive to population stratification which may easily lead to spurious association. Performing a joint analysis of LD and linkage mapping can help to overcome the limits of both approaches. Moreover, the advantages of the two complementary strategies can be utilized maximally. In practice, linkage analysis may be performed using pedigree data to identify suggestive linkage between markers and trait loci based on a sparse marker map. In the presence of linkage, joint LD and linkage mapping can be carried out to do fine gene mapping based on a dense genetic map using both pedigree and population data. Population and pedigree data of any type can be combined to perform a joint analysis of high resolution LD and linkage mapping of QTL by generalizing the method.