Sparse Selection of Multivariate Responses in Association Analysis: Study of the Effect of Air Particles on DNA Methylation in a Gene Set

• Activity Number: 138
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 8:30 AM to 10:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Tamar Sofer*+ and Arnab Maity and Brent Coull and Andrea Baccarellli and Joel Schwartz and Xihong Lin
• Companies: Harvard School of Public Health and North Carolina State University and Harvard School of Public Health and Harvard School of Public Health and Harvard School of Public Health and Harvard School of Public Health
• Keywords: Gene and Environment ; Epigenetics ; Sparsity ; Variable Selection ; Canonical Correlation Analysis ; Principal Component analysis

Abstract

In environmental epigenetics, one is interested in studying the effects of environmental exposures on DNA methylations in a genetic pathway, which often consists of a large number of genes and many of them are likely to be unaffected by exposures. We develop three Sparse Outcome Selection (SOS) methods for modeling the association between multivariate responses in a genetic pathway and exposures by selecting a subset of genetic outcomes in the pathway whose linear combination yields the highest correlation with a linear combination of exposure variables. We propose SOS Principal Components Analysis, which is a semi-supervised correlation method; SOS Canonical Correlation Analysis (CCA) and the step-forward CCA, which are supervised correlation methods. We investigate three criteria for selecting the tuning parameter, which include prediction correlation, Bayesian Information Criterion, and Correlation Information Criterion, which we developed. We compare the performance of these methods with existing methods via simulations, and apply the methods to the Normative Aging data to study the effects of exposure to airborne particulate matter on DNA methylation in the asthma pathway.