Summary of contributions to GAW Group 12: multivariate methods

Here we summarize the contributions to Group 12 of Genetic Analysis Workshop (GAW) 14, held in Noordwijkerhout, The Netherlands. The theme of this group, multivariate methods, covered a broad range of statistical applications. Most of the contributors considered Problem 1 of the GAW. However, one paper considered the bivariate analysis of two binary phenotypes generated by the simulated data in Problem 2. Some contributors focused on statistical issues involved in considering multiple variables, and others on extensions to the variance-components methodology for analysis of quantitative traits. Applications to the Collaborative Study on the Genetics of Alcoholism data identified a single-nucleotide polymorphism (SNP) on chromosome 4 associated with the ttth1-ttth4 phenotypes, and replicated previous findings of linkage on chromosome 4 for alcohol consumption, using microsatellite and SNP data.     

Summary of contributions to GAW Group 5: linkage mapping methods, Problem 2

Here I summarize the contributions to Group 5 of Genetic Analysis Workshop 14, held in Noordwijkerhout in The Netherlands. The theme of this group was linkage mapping methods applied to the simulated data (Problem 2). A variety of approaches were taken, and a number of questions were examined. In addition to testing for linkage in regions harboring known disease or modifying loci, or testing for linkage across the genome, the contributions addressed such issues as whether power/significance could be improved by making use of subphenotypes in addition to the primary disease phenotype, by modeling interactions between loci, or by using meta-analytic approaches to combine results from different populations. Most contributions were successful in identifying known disease loci D1 and D2, and for those contributions that examined the relevant regions, in identifying disease loci D3 and D4. The power to detect modifying loci D5 and D6 appeared to be lower. Some gain in power/significance was found from making use of subphenotypes and meta-analytic approaches.     

Heterogeneity: GAW Group 15

This paper summarizes the contributions to Group 15 of Genetic Analysis Workshop (GAW) 14, which focused on methods for dealing with heterogeneity in linkage and association analysis. A variety of methods were employed, ranging from manipulation of the phenotype and/or identification of endophenotypes prior to analysis, to statistical methods allowing for heterogeneity in the analysis of simple dichotomous phenotypes as provided with the data. Overall, it was difficult to draw broad conclusions from these applications. Groups that focused on the simulated data had fairly consistent success in mapping the major genes (but not the two minor genes). However, this appeared to be the case regardless of whether heterogeneity was explicitly modeled, either at the phenotypic or statistical level, with relatively crude statistical methods applied to the simulated simple dichotomous trait Kofendrerd personality disorder also enjoying considerable success.     

Using linkage and association to identify and model genetic effects: summary of GAW15 Group 4

Group 4 at Genetic Analysis Workshop 15 focused on methods that exploited both linkage and association information to map disease loci. All contributions considered the dichotomous trait of rheumatoid arthritis, using either affected sibpairs and/or unrelated controls. While one contribution investigated linkage and association approaches separately in genome-wide analyses, the remaining others focused on joint linkage and association methods in specific genomic regions. The latter contributions proposed new methods and/or examined existing methods that addressed whether one or more polymorphisms partially or fully explained a linkage signal, particularly the methods proposed by Li et al. that are implemented in the computer program Linkage and Association Modeling in Pedigrees (LAMP). Using simulated SNP data under linkage peaks, several contributions found that existing family-based association approaches such as those of Martin et al. and Lake et al. had power similar to LAMP and to several methods proposed by the contributors for testing that a single nucleotide polymorphism partially explains a linkage peak. In evaluating methods for identifying if a polymorphism or a set of polymorphisms fully accounted for a linkage signal, several contributions found that it was important to understand that these methods may be subject to low power in some situations and thus, a non-significant result was not necessarily indicative of the polymorphism(s) being fully responsible for the linkage signal. Finally, modeling the disease using association evidence conditional on linkage may improve understanding of the etiology of disease.     

Genetic association with rheumatoid arthritis-Genetic Analysis Workshop 15: summary of contributions from Group 2

The papers in presentation group 2 of Genetic Analysis Workshop 15 (GAW15) conducted association analyses of rheumatoid arthritis data. The analyses were carried out primarily in the data provided by the North American Rheumatoid Arthritis Consortium (NARAC). One group conducted analyses in the data provided by the Canadian Rheumatoid Arthritis Genetics Study (CRAGS). Analysis strategies included genome-wide scans, the examination of candidate genes, and investigations of a region of interest on chromosome 18q21. Most authors employed relatively new methods, proposed extensions of existing methods, or introduced completely novel methods for aspects of association analysis. There were several common observations; a group of papers using a variety of methods found stronger association, on chromosomes 6 and 18 and in candidate gene PTPN22 among women with early onset. Generally, models that considered haplotypes or multiple markers showed stronger evidence for association than did single marker analyses.     

Genetic association studies of alcoholism--problems with the candidate gene approach

In recent years, progress has been made in the identification of causative factors in most single gene disorders and those with genes of major effect. In comparison, no genes contributing to a complex disorder have been unambiguously identified. A number of reasons for this have been previously presented in theoretical papers. Alcoholism is such a complex illness and genetic studies into its underlying genetic causes have suffered from lack of power due to small subject numbers, poor selection of control subjects, and over-emphasis on markers with low prior probability of involvement.     

Genetic analysis of simulated oligogenic traits in nuclear and extended pedigrees: Summary of GAW10 contributions

Participants of GAW10 had available simulated data on phenotypic and marker data for 200 replicates of each of two different collections of pedigrees. The simulated phenotype was multivariate and oligogenic, and included a number of complexities. Participants took widely different approaches to analysis. We compare their results to identify analysis approaches and use of the data that had the greatest impact on the conclusions, accuracy of estimates, and power to identify genetic factors. © 1997 Wiley-Liss, Inc.     

Genetic analysis of Alzheimer's disease: A summary of contributions to GAW8

Participants in the Alzheimer's disease component of GAW8 had access to three collections of pedigrees, complete with marker data from chromosomes 19 and 21. There were a total of 94 independent pedigrees and more than 2,000 individuals. Onset of the disorder varied widely among pedigrees. These data are briefly summarized along with a discussion of the problems associated with performing genetic analyses of Alzheimer's disease. The majority of the workshop participants performed an analysis either with some of the data contributed to the workshop or with data simulated on pedigrees of the same structure and disease status as were contributed. There were also a few purely methodological contributions. The contributions are summarized in three general areas: family association and phenotype, linkage analysis, and heterogeneity tests. © 1993 Wiley-Liss, Inc.     

Alcoholism and related traits: a summary of Group 13 contributions

Ten groups set out to study the genetics of alcoholism, using various measures of alcohol dependence such as Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria, and related endophenotypes such as the electrophysiological evaluation of event-related potentials. The groups used both genome-wide microsatellite and single-nucleotide polymorphism (SNP) genotyping data in families selected from the Collaborative Study on the Genetics of Alcoholism. The majority of investigators studied alcohol-related phenotypes and chose linkage rather than association analysis. The analysis of SNP data presented several challenges, including marker linkage disequilibrium issues and computational limitations. Many groups pursued novel techniques, both in dealing with the SNP data and the definition of phenotypes. While there was a limited amount of concordance among linkage findings, it was very instructive to see so many new strategies at work. Generally the SNP genotype data seemed to yield more information for multipoint linkage analysis than the microsatellite data, a finding that will benefit the genetic analysis of complex disease in the future. A novel linkage peak was detected using the SNP markers.     

Genetic analysis of bipolar disorder: Summary of GAW10

Participants in the Bipolar Disorder component of Genetic Analysis Workshop 10 had access to five distributed data sets containing chromosome 18 marker data and five data sets containing chromosome 5 data. A total of 25 groups participated in analyses and applied a myriad of methodologically innovative approaches to these data. Contributors focused on how to: (1) best define the phenotype from the spectrum of affective diagnoses; (2) test for a parent-of-origin effect in the transmission of bipolar illness and assess whether sharing in affected sib pairs depends on the sex of the transmitting parent; (3) evaluate the effects of misspecification of marker allele frequencies; (4) examine the putative candidate loci provided; (5) investigate the mode of inheritance; and (6) perform a meta-analysis to combine multiple data sets in a single analysis. Taken as a whole, the results would appear suggestive, but not definitive for linkage to a bipolar susceptibility locus on chromosome 18. The evidence for linkage appeared to increase as the diagnostic definition of the phenotype was broadened. Multipoint analyses seem to provide less evidence. It is possible that, because adjacent markers may be present in different data sets, the multipoint methods are combining marker data from different studies in a more comprehensive way than single marker analyses. Evidence on chromosome 5 and evidence for candidate loci were minimal. A discussion of problems inherent in combined analyses is given. © 1997 Wiley-Liss, Inc.     

An oliogenic disease displaying weak marker assocations: A summary of contributions to problem 1 of GAW9

Problem 1 consisted of artificial data simulated from an oligogenic (four-locus) disease model. The data set contained 200 nuclear families with at least one affected child and 100 control families with no affected members. Two of the disease loci were intended to be detectable via association analysis but not necessarily via linkage analysis, and the other two were virtually undetectable in this data set. Participants used association analysis, linkage analysis, and segregation analysis to analyze these data. Their findings are summarized, and analysis strategies are discussed. ©1995 Wiley-Liss, Inc.     

Data mining and computationally intensive methods: summary of Group 7 contributions to Genetic Analysis Workshop 13

The Framingham Heart Study data, as well as a related simulated data set, were generously provided to the participants of the Genetic Analysis Workshop 13 in order that newly developed and emerging statistical methodologies could be tested on that well-characterized data set. The impetus driving the development of novel methods is to elucidate the contributions of genes, environment, and interactions between and among them, as well as to allow comparison between and validation of methods. The seven papers that comprise this group used data-mining methodologies (tree-based methods, neural networks, discriminant analysis, and Bayesian variable selection) in an attempt to identify the underlying genetics of cardiovascular disease and related traits in the presence of environmental and genetic covariates. Data-mining strategies are gaining popularity because they are extremely flexible and may have greater efficiency and potential in identifying the factors involved in complex disorders. While the methods grouped together here constitute a diverse collection, some papers asked similar questions with very different methods, while others used the same underlying methodology to ask very different questions. This paper briefly describes the data-mining methodologies applied to the Genetic Analysis Workshop 13 data sets and the results of those investigations.     

Summary of analyses of problem 2 simulated data for GAW11

The simulated data for Problem 2 of the 11th Genetic Analysis Workshop (GAW11) consisted of family linkage data and disease data from three populations, each population with different genetic parameters. The disease was simulated such that there were two genetically distinct diseases, one caused by a single locus with three alleles and the other by two epistatically interacting loci. The two diseases were clinically identical. Each of the diseases had a severe and mild form. One of the diseases was influenced by an environmental factor and one by an allele at one of the marker loci (association). Linkage data consisted of 300 polymorphic markers. Investigators analyzed these data in a myriad of ways, using most currently available linkage techniques and association analysis methods. In addition, investigators also tested newly developed methods, some of which took advantage of the existence of the environmental component. Other types of analyses included meta-analyses, methods of combining data from different studies, questions of replication, how much information is available in a single data set, and analyses for gene x gene and gene x environment interaction.     

Design considerations for association studies of candidate genes in families

In genetic epidemiologic studies, investigators often use generalized linear models to evaluate the relationships between a disease trait and covariates, such as one or more candidate genes or an environmental exposure. Recently, attention has turned to study designs that mandate the inclusion of family members in addition to a proband. Standard models for analysis assume independent observations, which is unlikely to be true for family data, and the usual standard errors for the regression parameter estimates may be too large or too small, depending on the distribution of the covariates within and between families. The consequences of familial correlation on the study efficiency can be measured by a design effect that is equivalent to the relative information in a sample of unrelated individuals compared to a sample of families with the same number of individuals. We examine design effects for studies in association, and illustrate how the design effect is influenced by the intra-familial distribution of covariate values such as would be expected for a candidate gene. Typical design effects for a candidate gene range between 1.1 and 2.4, depending on the size of the family and the amount of unexplained familial correlation. These values correspond to a modest 10% increase in the required sample size up to more than doubling the requirements. Design effect values are useful in study design to compare the efficiency of studies that sample families versus independent individuals and to determine sample size requirements that account for familial correlation.     

The role of haplotypes in candidate gene studies

Human geneticists working on systems for which it is possible to make a strong case for a set of candidate genes face the problem of whether it is necessary to consider the variation in those genes as phased haplotypes, or whether the one-SNP-at-a-time approach might perform as well. There are three reasons why the phased haplotype route should be an improvement. First, the protein products of the candidate genes occur in polypeptide chains whose folding and other properties may depend on particular combinations of amino acids. Second, population genetic principles show us that variation in populations is inherently structured into haplotypes. Third, the statistical power of association tests with phased data is likely to be improved because of the reduction in dimension. However, in reality it takes a great deal of extra work to obtain valid haplotype phase information, and inferred phase information may simply compound the errors. In addition, if the causal connection between SNPs and a phenotype is truly driven by just a single SNP, then the haplotype-based approach may perform worse than the one-SNP-at-a-time approach. Here we examine some of the factors that affect haplotype patterns in genes, how haplotypes may be inferred, and how haplotypes have been useful in the context of testing association between candidate genes and complex traits.     

Methods and applications of genetic analysis in the study of cardiovascular disease: A summary of contributions to GAW8

The analyses of data provided for the cardiovascular disease section of Genetic Analysis Workshop 8 are briefly summarized. Methods using twin, segregation, and linkage analyses are used to determine, characterize, and localize genetic components involved in the etiology of cardiovascular disease and to explore the interaction of these components with known environmental risk factors. © 1993 Wiley-Liss, Inc.     

Linkage analyses of rheumatoid arthritis and related quantitative phenotypes: the GAW15 experience

The group that formed on the theme of linkage analyses of rheumatoid arthritis RA and related phenotypes (Group 10) in the Genetic Analysis Workshop 15 comprised 18 sets of investigators. Two data sets were available: one was a real set provided by the North American Rheumatoid Arthritis Consortium and collaborators in Canada, France (European Consortium Of Rheumatoid Arthritis Families) and the UK; the other was a simulated data set modelled after the real data set. Whereas a majority of the investigators analyzed the RA affection status as a binary phenotype, a few contributions considered data on correlated quantitative traits such as anti-cyclic citrullinated peptide and rheumatoid factor-immunoglobulin M. The different investigators applied a wide spectrum of linkage methods. As expected, most methods could identify the human leukocyfeantigen region on chromosome 6 as a major genetic factor for RA. In addition, some novel chromosomal regions provided significant evidence of linkage in multiple contributions in the group. In this report, we discuss the different strategies explored by the different investigators with the common goal of improving the power to detect linkage.     

Testing association between disease and multiple SNPs in a candidate gene

Current technology allows investigators to obtain genotypes at multiple single nucleotide polymorphism (SNPs) within a candidate locus. Many approaches have been developed for using such data in a test of association with disease, ranging from genotype-based to haplotype-based tests. We develop a new approach that involves two basic steps. In the first step, we use principal components (PCs) analysis to compute combinations of SNPs that capture the underlying correlation structure within the locus. The second step uses the PCs directly in a test of disease association. The PC approach captures linkage-disequilibrium information within a candidate region, but does not require the difficult computing implicit in a haplotype analysis. We demonstrate by simulation that the PC approach is typically as or more powerful than both genotype- and haplotype-based approaches. We also analyze association between respiratory symptoms in children and four SNPs in the Glutathione-S-Transferase P1 locus, based on data from the Children's Health Study. We observe stronger evidence of an association using the PC approach (p = 0.044) than using either a genotype-based (p = 0.13) or haplotype-based (p = 0.052) approach.     

Exploring genetic analysis of complex traits through the paradigm of alcohol dependence: summary of GAW11 contributions

We discuss the Genetic Analysis Workshop 11 analyses of data from the Collaborative Study on the Genetics of Alcoholism from a methodological perspective, concentrating on approaches and issues relevant to linkage and association studies of complex human phenotypes. Genome screening by parametric linkage, nonparametric linkage, association, and combined linkage/association methods are discussed. Issues particular to complex disease include etiologic heterogeneity, multivariate phenotype modeling, and parent-of-origin effects. Other methodological topics discussed are new and enhanced methods, ascertainment, weighting of nonindependent sib pairs, and data cleaning and validation.