Tuesday, May 3 2022
2:30pm - 4:30pm
Masters Presentation
Evaluation of extending Proxy External Controls Association Test (ProxECAT) to Poisson Regression

Genome-wide association studies (GWAS) involve scanning complete sets of DNA, or genomes, in samples to identify genetic variations associated with particular diseases. Researchers have found that rare genetic variations can be indicators to having a pre-disposition for certain diseases and genetic conditions. Many statistical methods have been developed that can identify associations between genes and complex traits. In 2018, ProxECAT was developed for rare variant association studies, using case control data from internal and external sources gathered from publicly available databases. While providing a robust approach to rare variant association studies using external controls, ProxECAT has opportunities for improvement. ProxECAT cannot control for covariates or incorporate internal and external control data sets in the same statistical test. We explore these areas for improvement using Poisson regression and compare the results to ProxECAT and ProxECAT using logistic regression. We find that the implemented Poisson regression models cannot account for the imbalance of cases and controls in the data. We explore other possible avenues for continuing the evaluation of ProxECAT to improve on its limitations.
Speaker:Makayla Cowles
Location:Zoom link in email

Download as iCalendar