Lowering TGs to reduce the risk of coronary heart disease (CHD) is an active area of drug development. New drugs under development target TG genes in the lipoprotein lipase (LPL). The status quo is that: (1) we know little about potential beneficial and detrimental effects of long-term inhibition (or activation) of these target TG genes; (2) most TG-lowering drugs in development focus on the LPL pathway--we need to identify new TG targets in other pathways; (3) TG drug development currently targets one gene at a time and neglects agents that affect many genes simultaneously. We propose to fill the knowledge gaps as follows. (1) TG levels are associated with many diseases and TG genes regulate many biological processes; thus, long-term targeting of TG genes may have pleiotropic effects other than reducing CHD. Traditional post- marketing approaches to identify such effects require a long time. The effects of long-term inhibition of TG genes can be defined rapidly by studying individuals with genetically determined variation in gene function--a Mendelian randomization approach. In Aim 1 we will define clinical phenotypes other than CHD associated with genetically determined variation of TG gene function by using (a) known functional variants, (b) imputed gene expression, and (c) a gene-specific genetic risk score (GRS) as proxies of long-term effect of drugs targeting TG genes (LPL, APOC2, APOC3, ANGPTL3, and ANGPTL4) and testing their association with ~1,600 clinical phenotypes extracted from EHRs in BioVU (~130,000) and eMERGE (~100,000). (2) Identifying novel genes associated with TG levels will facilitate the development of TG-lowering drugs. The high genetic diversity in people of African ancestry (AAs) enhances our ability to identify variants with large effect size. A strategy of combining sequencing and extreme-tail sampling (studying people at the extremes of a quantitative trait) led to the development of PCSK9 inhibitors to lower LDL-C. In Aim 2, we will apply extreme- tail sampling and exome sequencing in AAs to identify new therapeutic targets for lowering TGs. (3) In addition to targeting one gene at a time, there is increasing interest in using the transcriptome for drug development by searching for drugs that reverse the transcriptomic signature associated with a disease. However, the measured transcriptome is affected by the disease itself and associated diseases and therapies. In contrast, the genetic component of the transcriptome is not confounded in this way and is more likely to represent a causal signal. In Aim 3, we will impute the genetically determined component of the TG transcriptome (i.e., the virtual transcriptome). By searching drug perturbation databases, we will identify repurposing drug candidates that reverse the TG virtual transcriptomic signature. The candidates identified will be validated by characterizing their effects on measured TGs in large EHRs (BioVU and eMERGE). These studies will have potential hig...