PROJECT SUMMARY / ABSTRACT The field of pharmacogenomics has progressed from the discovery of genetic variants that cause variable function of drug metabolism enzymes to clinical implementation of gene-guided drug prescribing. However, only a small number of drugs have clinically valid and actionable genetic associations. One problem with the current pharmacogenomic approach is a focus on genetic variants with a large effect on drug response among a small number of genes. For most drugs, the pathways of drug metabolism and response are complex. For these drugs, the effects of genetic variation on drug disposition and response are also likely to be complex, including effects of hundreds or thousands of genetic variants with variable effect size. The primary objective of this project is to quantitate and characterize the influence of variants throughout the genome on drug response outcomes. Using existing data sets, we will measure the impact of complex polygenic variation on response to a variety of drugs and drug classes. We will also explore rare genetic variation in fentanyl distribution. Aim 1 is to analyze the collective effect of all variants genotyped as part of prior genome-wide association studies for clopidogrel, statins, methotrexate, ACE-inhibitors, antidepressants, and vancomycin, in order to determine the genetic architecture for response to each drug. Through our analysis, we will use mixed models to quantitate the amount of variability in drug response can be attributed to all genetic variation captured using genome-wide genotyping. We will also measure the relative impact of variants with small, moderate, and large effect size. The findings from completion of Aim 1 will guide future efforts in pharmacogenomics. For drugs where nearly all genetic effects are mediated by a small number of well-established variants, the focus can shift from variant discovery to clinical implementation using the current paradigm of targeted genotyping. In contrast, for drugs with genetic effects due to hundreds of variants with variable effect size, validation of the polygenic models across diverse populations is the next step. Aim 2 applies the mixed models approach to the commonly used and highly variable drug, fentanyl. Using data from an ongoing fentanyl pharmacokinetic study, we will define the genomic architecture of fentanyl disposition in order to create a genomic predictor of fentanyl pharmacokinetics. The genomic predictor will then be validated in an independent dataset, providing the opportunity to test the clinical implementation of this genomic predictor in future research. In Aim 3 we will further explore fentanyl disposition, performing whole genome sequencing in individuals with highly atypical fentanyl drug concentrations in order to identify novel genes and rare variants driving fentanyl kinetics. Discovery of these new associations will illuminate biological mechanisms of fentanyl metabolism and transport. Overall, through a shift ...