Estimation and testing of gene-environment interactions in family-based association studies

Gene-environment interactions are of interest in genetic association studies for several reasons. First, the power to detect genetic effects may be substantially decreased if those effects differ according to environmental exposure and if no account is taken of this interaction with environmental exposure in the analysis. Second, such interactions may indicate a phenomenon of genuine biological interest (whereby a particular genetic effect operates only in the presence of an environmental trigger, or vice versa), understanding of which can lead us to a greater understanding of possible mechanisms and pathways in disease progression. Here I discuss the testing and estimation of gene-environment interactions via the case/pseudocontrol and related approaches. As originally proposed, the case/pseudocontrol approach applies to case/parents trios with no missing genotype data. I discuss some recent extensions that allow larger pedigree structures with some missing genotype data and present computer simulations to compare the performance of several competing approaches.     

Gene-environment interaction testing in family-based association studies with phenotypically ascertained samples: a causal inference approach

We propose a method for testing gene-environment (G × E) interactions on a complex trait in family-based studies in which a phenotypic ascertainment criterion has been imposed. This novel approach employs G-estimation, a semiparametric estimation technique from the causal inference literature, to avoid modeling of the association between the environmental exposure and the phenotype, to gain robustness against unmeasured confounding due to population substructure, and to acknowledge the ascertainment conditions. The proposed test allows for incomplete parental genotypes. It is compared by simulation studies to an analogous conditional likelihood-based approach and to the QBAT-I test, which also invokes the G-estimation principle but ignores ascertainment. We apply our approach to a study of chronic obstructive pulmonary disorder.     

A new powerful non-parametric two-stage approach for testing multiple phenotypes in family-based association studies

We introduce a new powerful nonparametric testing strategy for family-based association studies in which multiple quantitative traits are recorded and the phenotype with the strongest genetic component is not known prior to the analysis. In the first stage, using a population-based test based on the generalized estimating equation approach, we test all recorded phenotypes for association with the marker locus without biasing the nominal significance level of the later family-based analysis. In the second stage the phenotype with the smallest p value is selected and tested by a family-based association test for association with the marker locus. This strategy is robust against population admixture and stratification and does not require any adjustment for multiple testing. We demonstrate the advantages of this testing strategy over standard methodology in a simulation study. The practical importance of our testing strategy is illustrated by applications to the Childhood Asthma Management Program asthma data sets.     

Dealing with missing data in family-based association studies: a multiple imputation approach

To test for association between a disease and a set of linked markers, or to estimate relative risks of disease, several different methods have been developed. Many methods for family data require that individuals be genotyped at the full set of markers and that phase can be reconstructed. Individuals with missing data are excluded from the analysis. This can result in an important decrease in sample size and a loss of information. A possible solution to this problem is to use missing-data likelihood methods. We propose an alternative approach, namely the use of multiple imputation. Briefly, this method consists in estimating from the available data all possible phased genotypes and their respective posterior probabilities. These posterior probabilities are then used to generate replicate imputed data sets via a data augmentation algorithm. We performed simulations to test the efficiency of this approach for case/parent trio data and we found that the multiple imputation procedure generally gave unbiased parameter estimates with correct type 1 error and confidence interval coverage. Multiple imputation had some advantages over missing data likelihood methods with regards to ease of use and model flexibility. Multiple imputation methods represent promising tools in the search for disease susceptibility variants.     

Genomewide scan for anal atresia in swine identifies linkage and association with a chromosome region on Sus scrofa chromosome 1

Anal atresia is a rare and severe disorder in swine occurring with an incidence of 0.1-1.0%. A whole-genome scan based on affected half-sibs was performed to identify susceptibility loci for anal atresia. The analysis included 27 families with a total of 95 animals and 65 affected piglets among them. Animals were genotyped for 126 microsatellite markers distributed across the 18 autosomal porcine chromosomes and the X chromosome, covering an estimated 2080 cM. Single-point and multipoint nonparametric linkage scores were calculated using the computer package ALLEGRO 1.0. Significant linkage results were obtained for chromosomes 1, 3, and 12. Markers on these chromosomes and additionally on chromosomes for which candidate genes have been postulated in previous studies were subjected to the transmission disequilibrium test (TDT). The test statistic exceeded the genomewide significance level for adjacent markers SW1621 (P = 7 x 10(-7)) and SW1902 (P = 3 x 10(-3)) on chromosome 1, supporting the results of the linkage analysis. A specific haplotype associated with anal atresia that could prove useful for selection against the disorder was revealed. Suggestive linkage and association were also found for markers S0081 on chromosome 9 and SW957 on chromosome 12.     

Using the noninformative families in family-based association tests: a powerful new testing strategy

For genetic association studies with multiple phenotypes, we propose a new strategy for multiple testing with family-based association tests (FBATs). The strategy increases the power by both using all available family data and reducing the number of hypotheses tested while being robust against population admixture and stratification. By use of conditional power calculations, the approach screens all possible null hypotheses without biasing the nominal significance level, and it identifies the subset of phenotypes that has optimal power when tested for association by either univariate or multivariate FBATs. An application of our strategy to an asthma study shows the practical relevance of the proposed methodology. In simulation studies, we compare our testing strategy with standard methodology for family studies. Furthermore, the proposed principle of using all data without biasing the nominal significance in an analysis prior to the computation of the test statistic has broad and powerful applications in many areas of family-based association studies.     

Optimal unified approach for rare-variant association testing with application to small-sample case-control whole-exome sequencing studies

We propose in this paper a unified approach for testing the association between rare variants and phenotypes in sequencing association studies. This approach maximizes power by adaptively using the data to optimally combine the burden test and the nonburden sequence kernel association test (SKAT). Burden tests are more powerful when most variants in a region are causal and the effects are in the same direction, whereas SKAT is more powerful when a large fraction of the variants in a region are noncausal or the effects of causal variants are in different directions. The proposed unified test maintains the power in both scenarios. We show that the unified test corresponds to the optimal test in an extended family of SKAT tests, which we refer to as SKAT-O. The second goal of this paper is to develop a small-sample adjustment procedure for the proposed methods for the correction of conservative type I error rates of SKAT family tests when the trait of interest is dichotomous and the sample size is small. Both small-sample-adjusted SKAT and the optimal unified test (SKAT-O) are computationally efficient and can easily be applied to genome-wide sequencing association studies. We evaluate the finite sample performance of the proposed methods using extensive simulation studies and illustrate their application using the acute-lung-injury exome-sequencing data of the National Heart, Lung, and Blood Institute Exome Sequencing Project.     

On the meta-analysis of genome-wide association studies: a robust and efficient approach to combine population and family-based studies

For the meta-analysis of genome-wide association studies, we propose a new method to adjust for the population stratification and a linear mixed approach that combines family-based and unrelated samples. The proposed approach achieves similar power levels as a standard meta-analysis which combines the different test statistics or p values across studies. However, by virtue of its design, the proposed approach is robust against population admixture and stratification, and no adjustments for population admixture and stratification, even in unrelated samples, are required. Using simulation studies, we examine the power of the proposed method and compare it to standard approaches in the meta-analysis of genome-wide association studies. The practical features of the approach are illustrated with a meta-analysis of three genome-wide association studies for Alzheimer's disease. We identify three single nucleotide polymorphisms showing significant genome-wide association with affection status. Two single nucleotide polymorphisms are novel and will be verified in other populations in our follow-up study.     

Polygenic modeling of genome-wide association studies: an application to prostate and breast cancer

Genome-wide association studies (GWAS) have successfully detected and replicated associations with numerous diseases, including cancers of the prostate and breast. These findings are helping clarify the genomic basis of such diseases, but appear to explain little of disease heritability. This limitation might reflect the focus of conventional GWAS on a small set of the most statistically significant associations with disease. More information might be obtained by analyzing GWAS using a polygenic model, which allows for the possibility that thousands of genetic variants could impact disease. Furthermore, there may exist common polygenic effects between potentially related phenotypes (e.g., prostate and breast cancer). Here we present and apply a polygenic model to GWAS of prostate and breast cancer. Our results indicate that the polygenic model can explain an increasing--albeit low--amount of heritability for both of these cancers, even when excluding the most statistically significant associations. In addition, nonaggressive prostate cancer and breast cancer appear to share a common polygenic model, potentially reflecting a similar underlying biology. This supports the further development and application of polygenic models to genomic data.     

ChromoScan: a scan statistic application for identifying chromosomal regions in genomic studies

ChromoScan is an implementation of a genome-based scan statistic that detects genomic regions, which are statistically significant for targeted measurements, such as genetic associations with disease, gene expression profiles, DNA copy number variations, as well as other genome-based measurements. A Java graphic user interface (GUI) is provided to allow users to select appropriate data transformations and thresholds for defining the significant events. AVAILABILITY: ChromoScan is freely available from http://www.epidkardia.sph.umich.edu/software/chromoscan/     

Distribution of the number of false discoveries in large-scale family-based association testing with application to the association between <Emphasis Type="Italic">PTPN1</Emphasis> and hypertension and obesity

We present a model-free approach to the study of the number of false discoveries for large-scale simultaneous family-based association tests (FBATs) in which the set of discoveries is decided by applying a threshold to the test statistics. When the association between a set of markers in a candidate gene and a group of phenotypes is studied by a class of FBATs, we indicate that a joint null hypothesis distribution for these statistics can be obtained by the fundamental statistical method of conditioning on sufficient statistics for the null hypothesis. Based on the joint null distribution of these statistics, we can obtain the distribution of the number of false discoveries for the set of discoveries defined by a threshold; the size of this set is referred to as its tail count. Simulation studies are presented to demonstrate that the conditional, not the unconditional, distribution of the tail count is appropriate for the study of false discoveries. The usefulness of this approach is illustrated by re-examining the association between PTPN1 and a group of blood-pressure-related phenotypes reported by Olivier et al. (Hum Mol Genet 13:1885–1892, 2004); our results refine and reinforce this association.     

Comparison of multimarker logistic regression models, with application to a genomewide scan of schizophrenia

Background

Imprinted genes show expression from one parental allele only and are important for development and behaviour. This extreme mode of allelic imbalance has been described for approximately 56 human genes. Imprinting status is often disrupted in cancer and dysmorphic syndromes. More subtle variation of gene expression, that is not parent-of-origin specific, termed 'allele-specific gene expression' (ASE) is more common and may give rise to milder phenotypic differences. Using two allele-specific high-throughput technologies alongside bioinformatics predictions, normal term human placenta was screened to find new imprinted genes and to ascertain the extent of ASE in this tissue.

Results

Twenty-three family trios of placental cDNA, placental genomic DNA (gDNA) and gDNA from both parents were tested for 130 candidate genes with the Sequenom MassArray system. Six genes were found differentially expressed but none imprinted. The Illumina ASE BeadArray platform was then used to test 1536 SNPs in 932 genes. The array was enriched for the human orthologues of 124 mouse candidate genes from bioinformatics predictions and 10 human candidate imprinted genes from EST database mining. After quality control pruning, a total of 261 informative SNPs (214 genes) remained for analysis. Imprinting with maternal expression was demonstrated for the lymphocyte imprinted gene ZNF331 in human placenta. Two potential differentially methylated regions (DMRs) were found in the vicinity of ZNF331. None of the bioinformatically predicted candidates tested showed imprinting except for a skewed allelic expression in a parent-specific manner observed for PHACTR2, a neighbour of the imprinted PLAGL1 gene. ASE was detected for two or more individuals in 39 candidate genes (18%).

Conclusions

Both Sequenom and Illumina assays were sensitive enough to study imprinting and strong allelic bias. Previous bioinformatics approaches were not predictive of new imprinted genes in the human term placenta. ZNF331 is imprinted in human term placenta and might be a new ubiquitously imprinted gene, part of a primate-specific locus. Demonstration of partial imprinting of PHACTR2 calls for re-evaluation of the allelic pattern of expression for the PHACTR2-PLAGL1 locus. ASE was common in human term placenta.     

Genomic control, a new approach to genetic-based association studies

During the past decade, mutations affecting liability to human disease have been discovered at a phenomenal rate, and that rate is increasing. For the most part, however, those diseases have a relatively simple genetic basis. For diseases with a complex genetic and environmental basis, new approaches are needed to pave the way for more rapid discovery of genes affecting liability. One such approach exploits large, population-based samples and large-scale genotyping to evaluate disease/gene associations. A substantial drawback to such samples is the fact that population heterogeneity can induce spurious associations between genes and disease. We describe a method called genomic control (GC), which obviates many of the concerns about population substructure by using the features of the genomes present in the sample to correct for stratification. Two such approaches are now available. The GC approach exploits the fact that population substructure generate "overdispersion" of statistics used to assess association. By testing multiple polymorphisms throughout the genome, only some of which are pertinent to the disease of interest, the degree of overdispersion generated by population substructure can be estimated and taken into account. The other approach, called Structured Association (SA), assumes that the sampled population, while heterogeneous, is composed of subpopulations that are themselves homogeneous. By using multiple polymorphisms throughout the genome, SA probabilistically assigns sampled individuals to these latent subpopulations. We review in detail the overdispersion GC. In addition to outlining the published ideas on this method, we describe several extensions: quantitative trait studies and case-control studies with haplotypes and multiallelic markers. For each study design our goal is to achieve control similar to that obtained for a family-based study, but with the convenience found in a population-based design.     

Genomewide scan in Ashkenazi Jewish families demonstrates evidence of linkage of ocular refraction to a QTL on chromosome 1p36

The development of refractive error is mediated by both environmental and genetic factors. We performed regression-based quantitative trait locus (QTL) linkage analysis on Ashkenazi Jewish families to identify regions in the genome responsible for ocular refraction. We measured refractive error on individuals in 49 multi-generational American families of Ashkenazi Jewish descent. The average family size was 11.1 individuals and was composed of 2.7 generations. Recruitment criteria specified that each family contain at least two myopic members. The mean spherical equivalent refractive error in the sample was −3.46D (SD=3.29) and 87% of individuals were myopic. Microsatellite genotyping with 387 markers was performed on 411 individuals. We performed multipoint regression-based linkage analysis for ocular refraction and a log transformation of the trait using the statistical package Merlin-Regress. Empirical genomewide significance levels were estimated through gene-dropping simulations by generating random genotypes at each of the 387 markers in 200 replicates of our pedigrees. Maximum LOD scores of 9.5 for ocular refraction and 8.7 for log-transformed refraction (LTR) were observed at 49.1 cM on chromosome 1p36 between markers D1S552 and D1S1622. The empirical genomewide significance levels were P=0.065 for ocular refraction and P<0.005 for LTR, providing strong evidence for linkage of refraction to this locus. The inter-marker region containing the peak spans 11 Mb and contains approximately 189 genes. Conclusion: We found genomewide significant evidence for linkage of refractive error to a novel QTL on chromosome 1p36 in an Ashkenazi Jewish population.     

A composite likelihood approach to latent multivariate Gaussian modeling of SNP data with application to genetic association testing

Many statistical tests have been proposed for case-control data to detect disease association with multiple single nucleotide polymorphisms (SNPs) in linkage disequilibrium. The main reason for the existence of so many tests is that each test aims to detect one or two aspects of many possible distributional differences between cases and controls, largely due to the lack of a general and yet simple model for discrete genotype data. Here we propose a latent variable model to represent SNP data: the observed SNP data are assumed to be obtained by discretizing a latent multivariate Gaussian variate. Because the latent variate is multivariate Gaussian, its distribution is completely characterized by its mean vector and covariance matrix, in contrast to much more complex forms of a general distribution for discrete multivariate SNP data. We propose a composite likelihood approach for parameter estimation. A direct application of this latent variable model is to association testing with multiple SNPs in a candidate gene or region. In contrast to many existing tests that aim to detect only one or two aspects of many possible distributional differences of discrete SNP data, we can exclusively focus on testing the mean and covariance parameters of the latent Gaussian distributions for cases and controls. Our simulation results demonstrate potential power gains of the proposed approach over some existing methods.     

Closed testing with Globaltest,with application in metabolomics

The Globaltest is a powerful test for the global null hypothesis that there is no association between a group of features and a response of interest, which is popular in pathway testing in metabolomics. Evaluating multiple feature sets, however, requires multiple testing correction. In this paper, we propose a multiple testing method, based on closed testing, specifically designed for the Globaltest. The proposed method controls the familywise error rate simultaneously over all possible feature sets, and therefore allows post hoc inference, that is, the researcher may choose feature sets of interest after seeing the data without jeopardizing error control. To circumvent the exponential computation time of closed testing, we derive a novel shortcut that allows exact closed testing to be performed on the scale of metabolomics data. An R package ctgt is available on comprehensive R archive network for the implementation of the shortcut procedure, with applications on several real metabolomics data examples.     

A unified approach for quantifying, testing and correcting population stratification in case-control association studies

The HapMap project has given case-control association studies a unique opportunity to uncover the genetic basis of complex diseases. However, persistent issues in such studies remain the proper quantification of, testing for, and correction for population stratification (PS). In this paper, we present the first unified paradigm that addresses all three fundamental issues within one statistical framework. Our unified approach makes use of an omnibus quantity (delta), which can be estimated in a case-control study from suitable null loci. We show how this estimated value can be used to quantify PS, to statistically test for PS, and to correct for PS, all in the context of case-control studies. Moreover, we provide guidelines for interpreting values of delta in association studies (e.g., at alpha = 0.05, a delta of size 0.416 is small, a delta of size 0.653 is medium, and a delta of size 1.115 is large). A novel feature of our testing procedure is its ability to test for either strictly any PS or only 'practically important' PS. We also performed simulations to compare our correction procedure with Genomic Control (GC). Our results show that, unlike GC, it maintains good Type I error rates and power across all levels of PS.     

Development of an integrated 200K SNP genotyping array and application for genetic mapping,genome assembly improvement and genome wide association studies in pear (Pyrus)

Pear (Pyrus; 2n = 34), the third most important temperate fruit crop, has great nutritional and economic value. Despite the availability of many genomic resources in pear, it is challenging to genotype novel germplasm resources and breeding progeny in a timely and cost‐effective manner. Genotyping arrays can provide fast, efficient and high‐throughput genetic characterization of diverse germplasm, genetic mapping and breeding populations. We present here 200K AXIOM^® PyrSNP, a large‐scale single nucleotide polymorphism (SNP) genotyping array to facilitate genotyping of Pyrus species. A diverse panel of 113 re‐sequenced pear genotypes was used to discover SNPs to promote increased adoption of the array. A set of 188 diverse accessions and an F₁ population of 98 individuals from ‘Cuiguan’ × ‘Starkrimson’ was genotyped with the array to assess its effectiveness. A large majority of SNPs (166 335 or 83%) are of high quality. The high density and uniform distribution of the array SNPs facilitated prediction of centromeric regions on 17 pear chromosomes, and significantly improved the genome assembly from 75.5% to 81.4% based on genetic mapping. Identification of a gene associated with flowering time and candidate genes linked to size of fruit core via genome wide association studies showed the usefulness of the array in pear genetic research. The newly developed high‐density SNP array presents an important tool for rapid and high‐throughput genotyping in pear for genetic map construction, QTL identification and genomic selection.     

Floral and mating system divergence in secondary sympatry: testing an alternative hypothesis to reinforcement in Clarkia

Background and Aims

Reproductive character displacement (RCD) is often an important signature of reinforcement when partially cross-compatible taxa meet in secondary sympatry. In this study, floral evolution is examined during the Holocene range expansion of Clarkia xantiana subsp. parviflora from eastern Pleistocene refugia to a western zone of sympatry with its sister taxon, subsp. xantiana. Floral divergence between the two taxa is greater in sympatry than allopatry. The goal was to test an alternative hypothesis to reinforcement – that floral divergence of sympatric genotypes is simply a by-product of adaptation to pollination environments that differ between the allopatric and sympatric portions of the subspecies'' range.

Methods

Floral trait data from two common garden studies were used to examine floral divergence between sympatric and allopatric regions and among phylogeographically defined lineages. In natural populations of C. x. parviflora, the magnitude of pollen limitation and reproductive assurance were quantified across its west-to-east range. Potted sympatric and allopatric genotypes were also reciprocally translocated between geographical regions to distinguish between the effects of floral phenotype versus contrasting pollinator environments on reproductive ecology.

Key Results

Sympatric populations are considerably smaller flowered with reduced herkogamy. Pollen limitation and the reproductive assurance value of selfing are greater in sympatric than in allopatric populations. Most significantly, reciprocal translocation experiments showed these differences in reproductive ecology cannot be attributed to contrasting pollinator environments between the sympatric and allopatric regions, but instead reflect the effects of flower size on pollinator attraction.

Conclusions

Floral evolution occurred during the westward range expansion of parviflora, particularly in the zone of sympatry with xantiana. No evidence was found that strongly reduced flower size in sympatric parviflora (and RCD between parviflora and xantiana) is due to adaptation to limited pollinator availability. Rather, floral divergence appears to have been driven by other factors, such as interactions with congenerics in secondary sympatry.