Research
Effects of Genotyping on Brain Structure and Function
Through the MGH Brain Genomics Superstruct Project (GSP), in collaboration with Randy Buckner, PhD, Joshua Roffman, MD and colleagues, we have developed a large-scale resource for genomic studies of brain structure and function. This research has produced novel methods for estimating genetic contributions to high-dimensional brain imaging phenotypes as well identifying functional and structural circuits related to psychiatric conditions (anxiety, depression, schizophrenia) as well as normal variation.
Neural and Genetic Basis of Negative Valence Traits
Using state-of-the-art imaging and genomic approaches, this study aims to characterize the neural and genetic basis of anxiety-related phenotypes, including an indices of anxiety proneness involving the amygdala and medial pre-frontal cortex (mPFC). It also aims to characterize the the clinical and functional relevance of negative valence traits as defined by the NIMH RDoC (Research Domain Criteria). The development of RDoC by the NIMH is an effort to better define mental disorders through a ‘bottom-up’ approach that would incorporate multiple dimensions from the molecular to neurobiology to behavior.
Brain Genomics: The Neural and Genetic Architecture of Mental Illness (Morphometricity)
This study aims to develop a novel tool for prediction and prevention of psychiatric illness by estimating the proportion of observable variations in a trait that can be explained by brain morphology. Data collected as part of this study will be combined with existing data from the Mass General Scholars protocol to estimate the similarity between brain structure signatures, known as "cross disorder morphometricity."
The Development and Neural Bases of Emotion Processing
This collaboration with the laboratory of Charles A. Nelson, PhD and colleagues at Boston Children’s Hospital is examining the nature and neural architecture of fear perception, emotion and social processing the first seven years of life. Examining functional gene polymorphisms, genomewide genotyping, and telomere lengths polymorphism, we are evaluating the influence of genetic variation on individual differences in the development of emotion processing in children. This includes genetic analyses of both neurophysiologic and behavior as well as developmental trajectories of emotion processing.
Innovative Methods for Brain Imaging and Genomics
In efforts led by Tian Ge, we have been developing statistical, machine learning and computational methods to study the genetic underpinnings of brain structure and function. A significant aspect of our work is the application of heritability analyses to various measures of brain anatomy and physiology. Throughout this research, we developed the statistical approach MEGHA (massively expedited genome-wide heritability analysis), which made population-based heritability analysis of millions of phenotypes tractable for the first time. Using mixed effects modeling regression techniques, we also developed models for detecting gene-by-environment interactions and introduced a novel metric, termed “morphometricity,” as a measure of the neuroanatomical signature of a phenotype. Finally, we have been pioneering methods and applications to characterize the genetic architecture of human complex traits and improve disease risk prediction and stratification.
