Genetic epidemiologic analysis of quantitative phenotypes using Gibbs sampling

Gibbs sampling-based segregation analysis of asthma-associated quantitative traits in a population-based sample of nuclear families

Asthma is a common, complex human disease. Elevated serum immunoglobulin E (IgE) levels, elevated blood eosinophil counts, and increased airway responsiveness are physiological traits that are characteristic of asthma. Few studies have investigated major gene effects for these traits in a population-based sample. Further, it is not known if any putative major genes may be common to two or more of these traits. We investigated the existence and nature of major genes modulating asthma-associated quantitative traits in an Australian population-based sample of 210 Caucasian nuclear families. The sharing of these major genes was also investigated. Segregation analysis was based upon a Markov Chain Monte Carlo (Gibbs sampling) approach as implemented in the program BUGS v0.6. All models included adjustment for age, height, tobacco smoke exposure, and gender. The segregation of total IgE levels, blood eosinophil counts, and dose-response slope (DRS) of methacholine challenge were all consistent with major loci at which a recessive allele acted to increase or decrease the phenotype. The respective estimated frequencies of the recessive alleles were 68% (total IgE), 10% (blood eosinophil count), and 27% (DRS). Extensive modelling suggested that the major loci controlling total serum IgE levels, blood eosinophil counts, and airway responsiveness represent different genes. These data provide evidence, for the first time, of the existence of at least 3 distinct genetic pathways involving major gene effects on physiological traits closely associated with asthma. These results have implications for gene discovery programs.

Genetic variance components analysis for binary phenotypes using generalized linear mixed models (GLMMs) and Gibbs sampling

The common complex diseases such as asthma are an important focus of genetic research, and studies based on large numbers of simple pedigrees ascertained from population-based sampling frames are becoming commonplace. Many of the genetic and environmental factors causing these diseases are unknown and there is often a strong residual covariance between relatives even after all known determinants are taken into account. This must be modelled correctly whether scientific interest is focused on fixed effects, as in an association analysis, or on the covariances themselves. Analysis is straightforward for multivariate Normal phenotypes, but difficulties arise with other types of trait. Generalized linear mixed models (GLMMs) offer a potentially unifying approach to analysis for many classes of phenotype including multivariate Normal traits, binary traits, and censored survival times. Markov Chain Monte Carlo methods, including Gibbs sampling, provide a convenient framework within which such models may be fitted. In this paper, Bayesian inference Using Gibbs Sampling (a generic Gibbs sampler; BUGS) is used to fit GLMMs for multivariate Normal and binary phenotypes in nuclear families. BUGS is easy to use and readily available. We motivate a suitable model structure for Normal phenotypes and show how the model extends to binary traits. We discuss parameter interpretation and statistical inference and show how to circumvent a number of important theoretical and practical problems that we encountered. Using simulated data we show that model parameters seem consistent and appear unbiased in smaller data sets. We illustrate our methods using data from an ongoing cohort study.

Familial correlation of dietary intakes among postmenopausal women

A positive family history is a risk factor for many chronic diseases, including most cancers, coronary heart disease, and diabetes. Since diet is also associated with most chronic diseases, one possible explanation for non-Mendelian familial clustering is shared eating habits. Food frequency data were obtained on 3,515 sisters in the Iowa Women's Health Study, a prospective cohort of postmenopausal women. Intraclass correlations between sisters were computed on a range of energy-adjusted nutrients to determine whether dietary intakes were more similar among siblings than among unrelated individuals. Two methods were used to calculate correlations: analysis of variance modeling and weighted sibling correlations. F-tests and randomization tests were used to determine statistical significance. The intraclass correlations for all of the nutrients examined were statistically significantly greater than the hypothesized value of zero (P < 0.05). Representative correlations include dietary fiber (0.15), animal fat (0.12), vegetable fat (0.13), calcium (0.14), iron (0.04), cholesterol (0.08), sodium (0.10), vitamin D (0.16), and total energy intake (0.11). When corrected for measurement error, the magnitude of these correlations increased, on average 62%. Although modest in magnitude, these correlations may be high enough to influence familial clustering of complex diseases that are attributed, in part, to diet.

Analysis of gene-smoking interaction in lung cancer

The Louisiana Lung Cancer Dataset, consisting of 337 extended pedigrees, is analyzed to determine whether a major Mendelian gene interacts with cumulative tobacco smoking (pack-years). The proportional hazards model is utilized, as it is a natural framework for estimating relative risks while adjusting for variability in age of disease onset. Segregation analyses show evidence that a Mendelian gene is segregating in these families, with the most parsimonious model, including sex, pack-years, pack-years squared, and a dominant major gene. The estimated frequency of the high-risk allele is 2% and carriers are estimated to have relative risk of 17.3 for developing lung cancer, compared to noncarriers. The addition of a gene x pack-years interaction does not significantly improve the fit of the model, indicating that on a multiplicative scale, these two factors independently influence lung cancer risk. Smoking history is missing for 23% of the study subjects and degree of "missingness" depends on disease status, age, and birth-year. To account for the nonrandomness of the missing data, a Markov chain Monte Carlo method for covariate imputation is proposed and implemented. Results from this analysis also support a nonsignificant gene-smoking interaction and an allele frequency of 2%, but a lower genetic relative risk (9.0) compared to the "complete case" analysis.

Detection of gene-environment interactions in joint segregation and linkage analysis

We compare approaches for analysis of gene-environment (G x E) interaction, using segregation and joint segregation and linkage analyses of a quantitative trait. Analyses of triglyceride levels in a single large pedigree demonstrate the two methods and show evidence for a significant interaction (P=.015 when segregation analysis is used; P=.006 when joint analysis is used) between a codominant major gene and body-mass index. Genotype-specific correlation coefficients, between triglyceride levels and body-mass index, estimated from the joint model are rAA=.72, rAa=.49, and raa=. 20. Several simulation studies indicate that joint segregation and linkage analysis leads to less-biased and more-efficient estimates of a G x E-interaction effect, compared with segregation analysis alone. Depending on the heterozygosity of the marker locus and its proximity to the trait locus, we found joint analysis to be as much as 70% more efficient than segregation analysis, for estimation of a G x E-interaction effect. Over a variety of parameter combinations, joint analysis also led to moderate (5%-10%) increases in power to detect the interaction. On the basis of these results, we suggest the use of combined segregation and linkage analysis for improved estimation of G x E-interaction effects when the underlying trait gene is unmeasured.