The Correction of Length-Bias in Gene Set Analysis for DNA Methylation Data

• Activity Number: 193
• Type: Contributed
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Shaoyu Li* and Iwona Pawlikowska and Tong Lin
• Companies: The University of North Carolina at Charlotte and St. Jude Children's Research Hospital and St. Jude Children's Research Hospital
• Keywords: Kernel machine regression model ; length bias ; epigenome-wide association study (EWAS)

Abstract

Accompanied by the rapidly developing biotechnologies, genome-wide DNA methylation profiling has become feasible recently. New characteristics carried by the large scale DNA methylation data suggest existing statistical tools developed for other data type may not be directly suitable. In this work, we consider a novel 2-step statistical procedure aiming to resolve the length-bias issue in enrichment analysis for DNA methylation data. In the first step, the kernel machine regression method was applied to detect genes whose DNA methylation status are associated with phenotype. And a logistic regression model was used to identify enriched gene sets by incorporating gene size as a covariate to adjust its effect in the second step. Extensive simulation studies conducted shown the merits of the proposed produce comparing to some commonly used procedures. Finally, the proposed method was applied to a GEO DNA methylation data set to illustrate the feasibility of the proposed method.