Webcast: On the Replication of Genetic Associations: Timing Can Be Everything

Presenter: Jessica Lasky Su, Sc.C., Brigham and Women's Hospital

Date: April 15, 2008

Duration: 60 Minutes

ACCESS THIS WEBCAST:

Abstract

The failure to replicate genetic association findings is most commonly believed to be attributable to insufficient statistical power, population stratification, or various forms of between-study heterogeneity or environmental influences. Here, we illustrate another potential cause for non-replications that has so far not received much attention in the literature. We illustrate that the strength of a genetic effect can vary by age and that this phenomenon can cause “age-varying associations”. As we show by example, if not taken into account during the design and the analysis of the study, age-varying genetic associations can be the reason for non-replication. Using the 100K SNP scan of the Framingham Heart Study (FHS), we identify an age-varying association between a SNP in ROBO1 and obesity. Based on the data from the FHS, we hypothesized that there is an age-gene interaction. This finding was followed-up in eight independent samples comprised of 13,584 individuals. The association replicates in five of the eight studies, showing an age-dependent relationship (one-sided combined p = 3.92x10^-9, combined p-value from pediatric cohorts = 2.21x10-8, combined p-value from adult cohorts = 0.00422). Furthermore, this study illustrates that it will be difficult for cross-sectional study designs to detect age-varying associations. If the specifics of age/time-varying genetic effects are not considered in the selection of both the follow-up samples and in the statistical analysis, important genetic associations may be missed.

About the Presenter

Dr. Lasky Su is an Instructor in Medicine at Brigham and Women's Hospital. She has co-authored a number of publications on genetic association studies primarily focusing on family-based association testing. Dr. Lasky Su has also contributed several novel family-based methods implemented in the PBAT statistical package.