This project aims to develop and apply novel statistical approaches to address key challenges in the etiology of substance use, abuse and dependence. These are: high-density data; genotype by environment interplay; measurement; onset and offset; and comorbidity. Recent advances in genomic and other -omic technologies, neuroimaging, and near-continuous assessments of environment, location and behavior are creating unprecedented opportunities to understand the interplay between genetic and environmental risk factors for substance use behaviors. The project will develop new statistical models and computationally efficient software and methods to permit data mining of substance use phenotypes, genotypes and environmental measures. These models include those for: detecting GxE interaction in the presence of variable measurement precision; incorporating genetic marker and other high-density data into structural equation models to distinguish direct from indirect effects; models for gene-environment interplay, especially niche-selection; mixture distribution models of comorbidity; alternative models for symptoms co-occurrence; and regime switching to better characterize alternative pathways to outcomes. The new methods and models will be applied to datasets including measured genotypes and environmental risk factors collected in the United States, The Netherlands and Australia by project team members, The new models and methods will be disseminated freely and greatly increase the value of existing datasets and those being assembled with novel technologies.
Substance use abuse and dependence are especially complex, with genetic and environmental risk factors, with highly variable patterns of onset, offset and comorbidity with psychiatric disorders, cardiovascular disease and cancer. This project will develop statistical methods and software to improve the prediction of drug use, abuse and dependence, and its treatment and prevention.
Showing the most recent 10 out of 174 publications
