Semparametric Regression for Microarray Data using Support Vector Machines

• Activity Number: 62
• Type: Invited
• Date/Time: Monday, August 9, 2004 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Debashis Ghosh*+ and Xihong Lin and Dawei Liu
• Companies: University of Michigan and University of Michigan and University of Michigan
• Keywords: machine learning ; gene expression ; semiparametrics

Abstract

We consider a semiparametric regression model to relate a continuous outcome to clinical covariates and gene expressions, where the clinical covariate effects are modeled parametrically and gene expression effects are modeled nonparametrically using the support vector machine. The nonparametric function allows for the fact that the number of genes is likely to be large and the genes are likely to interact with each other. Equivalences with the linear mixed model will allow for the use of standard mixed model software. Both the regression coefficients of the clinical covariate effects and the support vector estimator of the nonparametric gene expression function can be obtained using the Best Linear Unbiased Predictor in linear mixed models. The smoothing parameter can be estimated as a variance component in linear mixed models. A score test is developed to test for the significant gene expression effects. The methods are illustrated using a prostate cancer dataset and evaluated using simulations.