Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 461 - Perils and Opportunities for Analyzing Biological, Behavioral, and Digital Phenotypes of Mental Functions
Type: Invited
Date/Time: Wednesday, August 10, 2022 : 2:00 PM to 3:50 PM
Sponsor: Mental Health Statistics Section
Abstract #320613
Title: Noisy BLUPS: Individual Prediction of Psychiatric Outcomes Using Imaging Data
Author(s): Wesley Thompson*
Companies: UCSD
Keywords: Mixed Models; High-Dimensional; Brain Imaging

Mixed-effects models have become essential for analyzing high-dimensional data from genome-wide association studies of complex traits. For example, Genome-Wide Complex Trait Analysis (GCTA) has become an essential tool to assess the total variance explained by genomic variation, which typically consists of many (hundreds or thousands) of causal variants spread across the genome, each with tiny effect but cumulatively explaining a substantial proportion of variation (the so-called "SNP heritability"). GCTA can also be used to compute Best Linear Unbiased Predictors (BLUPs) estimating the total genetic liability for a phenotype of interest. With the advent of large population-based studies, a similar phenomenon has occurred in neuroimaging studies of psychiatric outcomes: each unit of resolution (voxel or vertex) in a Magnetic Resonance Imaging (MRI) brain image explains a small proportion of behavioral and clinical outcomes, bit cumulatively images may explain more variation in total. Thus, researchers have begun borrowing tools such as GCTA and applying them to imaging studies to estimate variance explained and brain-based liability for behaviors such as psychiatric outcomes. Howev

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

Back to the full JSM 2022 program