PROJECT ABSTRACT The Cole Lab’s overarching vision is to harness unique characteristics of the human genome to gain deeper insights into the impact of environmental risk factors on human disease. Our research focuses on two key areas. First, we aim to develop methods that address gaps in the derivation, analysis, and interpretation of phenotypes and their associations with markers of human health. Second, we utilize these and other approaches to identify risk factors and mechanisms related to disease, consequently driving the evolution of novel methods. These two pillars of my lab are synergistic, continually reinforce one another, and will support my lab’s sustainable growth. Over the next five years, our primary goals are to develop methods and expand resources for the broader scientific community that improve the derivation and utilization of environmental phenotypes for human disease research. One, we will improve the curation and optimization of phenotype pre-processing decisions using genetic heritability as an unbiased metric for phenotype precision. Two, we will use genetic correlation as a proxy for phenotypic correlation in non-overlapping individuals to evaluate phenotype similarity and harmonize traits across diverse groups of individuals. Third, to improve the accurate identification and interpretation of the role environmental risk factors play in human disease, we will extend traditional genetic causal inference methods (i.e. Mendelian randomization) to better handle excess cross-trait correlation and genetic instrument mis- specification commonly seen with environmental risk factors. Furthermore, we will use comprehensive sets of phenotypes in diverse, large-scale biobanks (UK Biobank and AllofUs) to develop and optimize these methods, and will make the methods and results publicly available for a broad range of scientists to utilize. Both in the short-term and long-term, we aim to expand public datasets with more comprehensive, meticulously curated phenotypes. In the next five years, we will link clinical trials with high-quality longitudinal phenotypes at the University of Colorado (CU) with data available through the Colorado Biobank, including genetics, health records, surveys, and geocoded data. This will enhance this resource for CU investigators, and specifically enable our group to investigate the effects of genetic risk on intervention response and clinical-grade phenotypes. Our efforts in the next five years will support our research focus on applied statistical genetics to identify risk factors and mechanisms related to disease, with an emphasis on metabolic disease as a model system for which our group has expertise. In the long-term, this work allows our group to identify novel research gaps for new method development. Furthermore, our applied research positions our team to best identify the challenges of uniting scientists across disciplines and skillsets and the gaps in current resources. Together this research and ...