Statistical, population genetics and genetic epidemiology

点击翻译按钮获取中文摘要

Our research have been in the area of development and application of statistical approaches for gene mapping and multidimensional data. We have been researching functional linear models for genetic association studies and continued to work on ascertaining the proportion of real and spurious signals among top hits in genomic studies with many tests. Recent technological advances equipped researchers with capabilities that go beyond traditional genotyping of loci known to be polymorphic in a general population. Genetic sequences of study participants can now be assessed directly. This capability removed technology-driven bias toward scoring predominantly common polymorphisms and let researchers reveal a wealth of rare and sample-specific variants. While the relative contributions of rare and common polymorphisms to trait variation are being debated, researchers are faced with the need for new statistical tools for simultaneous evaluation of all variants within a region. Several research groups demonstrated flexibility and good statistical power of the functional linear model approach. We have been extending previous developments to allow inclusion of multiple traits and to provide capability to do statistical adjustment for additional covariates. Our functional approach is unique in that it provides a nuanced depiction of effects and interactions for the variables in the model by representing them as curves varying over a genetic region. Our statistical research demonstrated flexibility and competitive power of our proposed approach by contrasting its performance with commonly used statistical tools. In collaboration with Dr. Diatchenko (McGill University) we explored applications of this approach for uncovering genetic architecture of genetic risk factors involved in the development of chronic pain conditions. In studies of relative contribution of an individual's genetic composition to the perception of pain, the general characteristics of pain sensitivity are typically measured by a wide range of different, yet possibly related pain phenotypes. Testing each of these pain-perception traits individually is subject to problems of multiple testing and may result in low statistical power. Furthermore, pain-related traits may share common etiology. Our approach allowed both simultaneous testing of multiple correlated phenotypes, including quantitative, binary, categorical, with adjustment for additional covariates. Another line of our research is on estimation of proportion of spurious findings among most statistically significant results. This topic is related to concerns about low replicability of scientific findings, which is in part related to misapplications of statistical analysis. Measures of statistical significance (P-values) are commonly used. Attempts to design simple ways to convert an association P-value into the probability that a finding is spurious have been met with difficulties. In our research, we proposed a method that lets researchers extract probability that a finding is spurious directly from a P-value.