Biostatistics Dissertation Defense

Yi Liu: Novel Test Procedures for Large-scale Bivariate Time-to-event Data on Single and...

Friday 7/28 1:00PM - 3:00PM
Public Health 7139

Yi Liu of the Department of Biostatistics defends her dissertation on "Novel Test Procedures for Large-scale Bivariate Time-to-event Data on Single and Gene-based Markers, with an Application to a Genetic Study of AMD Progression"

Graduate faculty of the University and all other interested parties are invited to attend


Motivated by a genome-wide association study (GWAS) to discover risk variants for the progression of Age-related Macular degeneration (AMD), I develop a computationally efficient copula-based score test, in which the association between bivariate progression times is explicitly modeled. Specifically, a two-step estimation approach with numerical derivatives to approximate the score function and information matrix is proposed. Both parametric and weakly parametric marginal distributions under the proportional hazards assumption are considered. Further I extend this work to gene-based tests through the functional linear smoothing approach, which models the variants (within the same gene region) as a function of their physical positions. A robust variance approach under functional linear model framework for bivariate variance adjustment is also been proposed. In both works, simulation studies were conducted to evaluate the Type I error control and power performance of the proposed method. Finally, we apply our method on a large randomized trial data, Age-related Eye Disease Study (AREDS), to identify susceptible risk variants and gene regions for AMD progression. The top variants identified in CFH and ARMS2 gene on Chromosome 1 and 10 show differential progression profiles for different genetic groups, which are useful in characterizing and predicting the risk of progression for patients with moderate AMD.

Last Updated On Monday, October 23, 2017 by Valenti, Renee Nerozzi
Created On Friday, June 09, 2017