Building Bridges to Allow Cross-species Translational genetics for the Study of Addiction

PROJECT SUMMARY Substance use disorders are among the most common psychiatric disorders and are a leading cause of disability throughout the world. Over the last decade, genome-wide association studies (GWAS) have identified numerous genetic loci that contribute to addiction (opioid, tobacco, cannabis, alcohol). Turning these discoveries into mechanistic insights, a necessary first step for understanding the pathophysiology of addiction, and ultimately leading to the development of new therapies, is challenging. Model organisms serve as excellent tools to understand how human genomic variation affects traits. However, integration of human GWAS data with studies in model organisms has been limited. This is because GWAS do not lead to the identification of genes, but genetic variants (or SNPs), which cannot be easily translated across species. In addition, human GWAS have shown that risk for addiction is highly polygenic, but the existing strategies for cross-species translation do not capture the polygenic architecture of addiction. I am proposing an innovative solution to this problem by developing a framework for cross-species polygenic translation. Polygenic risk scores (PRS), which are a widely used tool for human genetic studies, predict risk for a trait by summing the contributions of numerous SNPs. Because these SNPs are species-specific, it is not possible to apply a PRS to another species. I am proposing to use transcriptomic analyses to overcome this obstacle. Leveraging resources and techniques from well-established statistical genetic tools, I will develop a method that will allow translation of polygenic signals from humans to rodents, and vice- versa. This will be accomplished by using the following steps: 1) use GWAS for addiction-related traits in human and model organisms to compile catalogs of genetic variants; 2) use transcriptomic data from GTEx and analogous datasets from model organisms to build gene prediction models, allowing estimation of transcript levels in individuals based on genotype information, 3) determine the association between these estimated transcript levels and addiction-related traits and 4) use these gene-level associations (rather than SNP level associations) to calculate Polygenic Transcriptomic Risk Scores (PTRS). This approach translates SNPs into gene estimated gene expression levels and then takes advantage of the fact that, while the same SNPs do not exist across species, gene orthology can be used to translate between species. In addition, I will use complementary methodologies, including cross-species network analyses and other tools, that also account for the polygenic nature of addiction. Developing tools that allow polygenic studies to share information between humans and model organisms will be transformative by opening up entirely new lines of research. For example, PTRS will provide a novel means of validating animal models of addiction, as it will be possible to empirically test whether the genetic signature for addiction in humans is related to the genetic signature of addiction phenotypes in rodents. PTRS will also serve as a toolkit for drug repositioning, namely studies aimed at identifying small molecules and other interventions that can alter the global gene expression in model organisms in a way that lower risk, as predicted by PTRS and network- based analyses.