Work has progressed in family-based statistical methods for studying genetic effects on qualitative and quantitative traits. We had previously developed methods for qualitative traits using genotypes for affected individuals and their parents, and if possible their grandparents. The log-linear approach we had proposed provides a powerful likelihood ratio test and estimation of relative penetrances for variant alleles, without requiring knowledge of the genetic model and it is robust against bias due to population stratification. The method can incorporate parent-of-origin effects, and maternally-mediated genetic effects, and allows for the possibility that one or both parental genotypes, or even the genotype of the affected offspring, may be missing. ? ? The assessment of transmission distortion from grandparents to affected grandchildren was shown to be much more powerful than methods based on transmissions from parents and this year the grandparents approach was applied to a study of ALL leukemia carried out in Montreal, where we found evidence for distortion in transmission of the NQ01 C609T allele. A variant of the design, called the 'pent' design because 5 individuals are genotyped for each case, was proposed, where cases, parents, and the maternal grandparents are studied. This approach provides almost as much statistical power as the full grandparents design. ? ? We have now extended our general approach using triads to allow mapping genes related to a quantitative trait, and we showed that our method provides improved power and robustness compared to analytic alternatives in widespread use. We also have extended the method to allow for multiple offspring from the same nuclear family. ? ? For early onset disorders, or traits such as birth weight or childhood adiposity, the prenatal environment can be particularly important and maternally-mediated genetic effects expressed during gestation can influence the phenotype of the offspring. An extension of our method for quantitative trait analysis allows the identification of maternally-mediated genetic effects and also for variant imprinted genes, where the expression of a single inherited copy of a variant allele may depend on the parent of origin. We have also extended our methods to allow for gene-by-environment interaction. Statistical methods for assessing maternal genetic effects on a qualitative outcome have been written in a chapter, for inclusion in a book related to maternal genetic effects being edited by Laura Mitchell.? ? Family-based studies cannot directly assess effects of exposures, which is a serious limitation; traditional case-control studies have different limitations. For example, they are subject to bias due to selection of controls and they cannot directly assess maternally mediated effects or imprinting. We developed a hybrid design that includes both nuclear families and unrelated population controls, and thereby offers the best of both approaches. This hybrid design is considerably more powerful than either traditional approach and also provides a test for bias due to population stratification, a problem that can invalidate a population-based case-control study of genetic effects. If this problem is detected in a given hybrid study, the investigator retains the ability to fall back to the case-parents component of the study, which will remain valid for testing genetic effects and gene-by-environment interaction. This design is being implemented in an ongoing study of the birth defects oral clefting. In work presented at the Joint Statistical Meeting this August, we extended the hybrid analysis to allow for gene by exposure interactions and showed that also for studying interactions the hybrid method is considerably more powerful than either a triad design or a case-control design with the same number of cases. ? ? We (Weinberg, Shi, and Umbach) have recently been developing data mining approaches for analysis of dense SNP scans in nuclear families. Our goals in this are fourfold: 1. to devise a method that does not rely on knowing or statistically estimating unknown haplotype phases but only uses the genotypes directly; 2. to develop a method that would produce a single statistical test when many tightly linked SNPs are studied in a gene or region; 3. to develop a method that would tolerate the typical swiss-cheese missing data phenomenon, where many individual SNPs are sporadically missing; and 4. to develop a method that would allow the data to nominate a causative haplotype, if one exists. We carried out realistic simulations of our approach by simulating a population based on haplotype frequencies for haplotypes for four genes derived from the Hapmap project. In constructing the simulations we assumed that each haplotype in turn carried a causative SNP mutation half of the time and that SNP either has (for some analyses) or has not (for others) been measured. Having demonstrated reasonably good operating characteristics for our method, we applied it to data on the birth defect, oral clefting and identified a set of 18 risk-tagging SNPs for a causative haplotype for clefting. The method extends to identifying haplotypes with maternally-mediated effects on risks and an extension of the method can be used to assess imprinting. Our haplotype paper is almost ready for submission.
Showing the most recent 10 out of 42 publications
