Biostatistics Seminar Series

Jenna Colavincenzo Carlson - Methods for family-based designs in genetic studies of orofacial clefti

Thursday 3/17 3:30PM - 4:30PM
Public Health Lecture Hall (A115)
Jenna Colavincenzo Carlson University of Pittsburgh, Department of Biostatistics Methods for family-based designs in genetic studies of orofacial clefting Orofacial clefts (OFCs) are a common, complex birth defect, affecting approximately 1/800 live births worldwide. Over the past decade, genetic studies have identified many regions associated with OFCs, but the total variation observed in clefting remains largely unexplained. Current statistical methods to identify risk variants for OFCs primarily use either case-control or family data. We present a summary of methods to jointly analyze case-control and family data. Some of these methods make further assumptions about the data including the assumption of Hardy-Weinberg equilibrium (HWE). Assuming HWE generates more efficient estimators; however, even modest deviation from HWE leads to inflation of type-I error. Methods not assuming HWE are obviously robust in this situation, but are less efficient. We propose a new method which gains power by assuming HWE for the underlying population yet is resistant to bias when that assumption is violated. This method estimates disease parameters by “shrinking” an unconstrained estimator toward a constrained estimator through an empirical-Bayes-type shrinkage approach. 

Last Updated On Tuesday, March 22, 2016 by Deleted User
Created On Tuesday, March 22, 2016