Online Program Home
  My Program

Abstract Details

Activity Number: 237 - Feature Selection and Statistical Learning in Genomics
Type: Contributed
Date/Time: Monday, July 31, 2017 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistics in Genomics and Genetics
Abstract #322941
Title: Graph Constrained Regularization for Nonparametric Instrumental Variable Regression in Genetical Genomic Analysis
Author(s): Bin Gao* and Yuehua Cui
Companies: Janssen Research & Development, LLC and Michigan State University
Keywords: Graph constrained regularization ; Gene selection ; Instrumental variable regression ; Regulatory network
Abstract:

Gene regulatory networks contain abundant information about the functions of gene expressions and the mechanisms of gene regulations. Thus, incorporating network structures would potentially increase the accuracy of gene selection and phenotype prediction. In this work, we used gene expressions to predict phenotypic responses while considering the graphical structures on gene networks. The model we used is an instrumental variable regression model with varying coefficients. We treated genetic variants as instrumental variables to deal with the endogeneity issue and applied varying coefficients to capture potential nonlinear environmental modulation effects. We proposed a two-step estimation procedure. In the first step, we applied Lasso to estimate the effects of genetic variants. In the second step, we used the predicted expressions obtained from the first step as predictors while adopting a network constrained regularization method to improve the efficiency of gene selection and estimation. Selection consistency is established under some assumptions. Simulation and real data analysis were conducted to demonstrate the effectiveness of our method compared to its counterpart.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association