Alcohol consumption is an etiologic essential for the development of alcohol use disorder and many alcohol pathologies. Despite evidence that genetic variations contribute to alcohol consumption, the genetic architecture remains poorly characterized. In terms of the genetic component, individual susceptibility to complex traits such as alcohol consumption arise mainly from variation in gene regulation. It has become increasingly clear in recent years that alternative polyadenylation (APA) ? a mechanism by which a single gene encodes multiple RNA isoforms with different polyadenylation (polyA) sites ? is extensively used to modulate gene regulation. An improved understanding of the genetic causes of predisposition to alcohol-related problems will facilitate a precision medicine approach through better prevention, diagnosis, and treatment. Yet shortcomings in current methods for concurrent APA identification and RNA transcript quantitation in high-throughput short read RNA sequencing (RNA-Seq) studies have limited our ability to study its genetic contributions to alcohol-related phenotypes. To remedy this, a novel trans-omics method and related software will be developed here that can precisely identify expressed polyA sites and accurately quantify individual isoforms including APA transcripts from a sequenced RNA-Seq library. Specifically, a supervised machine learning algorithm will be built that utilizes both RNA-Seq data and DNA sequence indicators to predict the locations of expressed polyA sites (Aim 1). The ?use-all-data? approach will identify and quantify APA isoforms present in the data with greater precision than current, one-dimensional omics techniques. Furthermore, the algorithm will be designed for integration into existing bioinformatics pipelines for quantitation along with other RNA species (Aim 2). To determine the influence of APA on the genetic predisposition to alcohol consumption, a systems genetics approach will be applied to relate brain RNA expression of the APA isoforms and other RNA molecules, including protein-coding, long non-coding, and microRNA (miRNA) transcripts, to alcohol consumption measures in a two-bottle choice paradigm (Aim 3). From this approach, if and how APA influences gene networks and susceptibility to miRNA regulation, what genetic factors control APA, and how APA may influence alcohol consumption can be evaluated. An extensive fellowship training plan and a team of mentors with diverse skillsets will be utilized to accomplish the multidisciplinary research project. The result of this training will be an alcohol researcher with not only the latest bioinformatics and computational skills but also the knowledge and experience to interpret, refine, and improve these models in the context of alcohol research. By following a training plan that incorporates the best of both disciplines, this proposal will be an example of a new type of alcohol researcher that can harness big data in a disease relevant way.
Understanding the role of alternative polyadenylation (APA) in alcohol-related phenotypes and pathologies will provide important insights into biological mechanisms related to alcohol for improved diagnosis, treatment, and prevention of alcohol-related problems. The proposed research will study the association of APA with a predisposition for free-choice alcohol consumption using a novel trans-omics computational approach.
