sahir.bhatnagar

Strong Heredity Penalized Regression Models for Non-Linear Gene-Environment Interactions

Diseases are now thought to be the result of changes in entire biological networks whose states are affected by a complex interaction of genetic and environmental factors. In general, power to estimate interactions is low, the number of possible interactions could be enormous and their effects may be non-linear. Existing approaches such as the lasso might keep an interaction but remove a main effect, which is problematic for interpretation. We develop a model for linear and non-linear interactions in penalized regression models that automatically enforces the strong heredity property. A computationally efficient fitting algorithm combined with a non-parametric screening approach scales to high-dimensional datasets and has been implemented in an R package. We apply our method to identify gene-prenatal maternal depression interactions on negative emotionality in mother–infant dyads from the Maternal Adversity, Vulnerability, and Neurodevelopment (MAVAN) cohort.

Date and Time: 

Tuesday, June 13, 2017 -
11:20 to 11:35

Co-authors (not including you): 

Yi Yang
McGill University
Alexia Jolicoeur-Martineau
Jewish General Hospital
Ashley Wazana
McGill University
Celia Greenwood
Lady Davis Institute

Language of Oral Presentation: 

English

Language of Visual Aids: 

English

Type of Presentation: 

Contributed

Session: 

Speaker

First Name Middle Name Last Name Primary Affiliation
edit Sahir R Bhatnagar McGill