BiographyDr. Thompson earned his Ph.D. in Statistics from Rutgers University in 2003, with a focus on statistical methods for longitudinal data analysis. He was appointed Assistant Professor of Statistics and Psychiatry at the University of Pittsburgh in 2005, where he received a five year NIH K25 Career Development Award to develop novel methods for studying co-variation in brain function and depression. Dr. Thompson joined the UCSD Department of Psychiatry in 2008. His current work involves Bayesian semi-parametric and mixture models with applications to (i) improving effect size estimation, replication, and prediction in genome-wide association studies, (ii) predicting onset of illness from multivariate biomarker trajectories, (iii) applications of functional data analysis to functional MRI data.
Research InterestsDr. Thompson's research focuses on the development and application of semi-parametric Bayesian hierarchical and mixture models for multivariate data, with applications to diverse areas of biological psychiatry. These applications include: 1) human brain imaging and other biological markers of mental illness; 2) improved replication and polygenic risk score estimation from genome-wide association studies; 3) estimation of developmental trajectories and prediction of outcomes from incomplete longitudinal data; and 4) prediction of clinical outcomes from intensively sampled longitudinal data.