Evaluation of Genetic Risk Score Models in the Presence of Interaction and Linkage Disequilibrium

• Activity Number: 311
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Ronglin Che*+ and Alison Motsinger-Reif
• Companies: North Carolina State University and North Carolina State University
• Keywords: explained variance ; genetic risk score (GRS) ; heritability ; interaction ; linkage disequilibrium (LD) ; predictive modeling

Abstract

Risk predictive modeling is becoming an important area of translational success in human genetics. The genetic risk score (GRS) modeling using multiple variants has been widely applied in real data, but not extensively studied from a methodological point of view. Previously, we compared the performance of a simple, additive GRS with our weighted GRS under simple assumptions without accounting for many of the complications found in real studies. In the present study, we employed various strategies to simulate a range of models demonstrating statistical interaction and linkage disequilibrium (LD) (indirect mapping). Three GRS models were compared in terms of power, type I error, AUC and AIC, including a simple count GRS (SC-GRS), an odds ratio weighted GRS (OR-GRS) and an explained variance weighted GRS (EV-GRS). Simulation factors included allele frequency, effect size, strength of interaction, degree of LD and heritability. We extensively examined the extent to how these interactions could influence the performance of models. Results show that the weighted methods outperform simple count method in general even if interaction or LD is present, with well controlled type I error.