2016-11-11

Olga A. Vsevolozhskaya , University of Kentucky

"Characterizing uncertainty in genetic association landscapes by functional Bayesian bands: Application to pleiotropic effects of CSF levels of Alzheimer’s disease proteins"

Cerebrospinal fluid (CSF) analytes harbor potential as diagnostic biomarkers for Alzheimer’s Disease (AD). Quantitative measures of CSF proteins comprise a set of often highly correlated endophenotypes that have previously shown promise in genetic analyses (Cruchaga et al., 2013; Kauwe et al., 2014). Pleiotropic impact of genetic variations on this set may provide additional insights into AD pathology at its earliest stages. To determine which specific endophenotypes are pleiotropic, one can employ methods based on the reverse regression of genotype on phenotypes. Recently, we proposed a method based functional linear models that utilizes reverse regression and simultaneously evaluates all variants within a genetic region for an association with multiple correlated phenotypes. Although the approach has the potential to answer lots of interesting questions, e.g., help estimate pleiotropic effects or the so¬called ‘treatment¬by¬trait’ interaction, it is subject to new challenges. For instance, to infer statistical significance of a genetic effect, one may construct pointwise confidence bands at each variant’s positions, which may be subject to multiple testing issue. To overcome thеsе issues, we extend our functional approach to Bayesian framework, which provides a statistical solution immune to both the problem of multiple testing and the winner’s curse. We show the validity of our approach through a simulation study and utilize it to explore pleiotropic effects of CSF analtyes based on the Alzheimer's Disease Neuroimaging Initiative (ADNI) data.

Stay connected TwitterFacebookLinkedInYouTubeInstagram