Alzheimer's Disease and Related Dementias (AD/ADRD) are common progressive conditions that damage critical mental functions, causing significant emotional, physical, and financial burdens. Currently, there is no cure for AD/ADRD, nor a drug to delay the onset of the disease. New drugs targeting AD/ADRD have had a 99% failure rate due to the incompleteness of biological knowledge, the emphasis on testing single therapies, the lack of predictive validity in animal models, and the emergence of unacceptable side effects. Drug repurposing--identifying new uses for existing drugs--offers an alternative approach that can reduce the time, costs, and risks of failure associated with new drug development. Indeed, genetic research and large biobanks have accumulated a wealth of relevant data for AD/ADRD drug repurposing efforts. However, given the complexity of AD/ADRD pathogenesis, identifying and prioritizing drug repurposing candidates for downstream analyses remain a challenge. Standard repurposing efforts generally arise from basic science or clinical observations, resulting in a one-at-a-time approach via candidate validation in clinical trials. Such efforts ignore the opportunities provided by big data (e.g., genetics and electronic health records) for identifying and validating candidates. The lack of standardized approaches to systematically screen and validate drug repurposing signals represents a critical barrier to the success of leveraging existing public knowledge and clinical data to advance AD/ADRD treatment. We propose integrating genetics, transcriptomics, literature, and clinical data with advanced informatic technologies and computational capabilities to develop and share tools which systematically screen, prioritize, and validate drug repurposing candidates and their combinations for AD/ADRD. The proposed project directly responds to PAR-20-156’s challenge to "combine large-scale genotypic and phenotypic data, and use established or develop novel computation approaches to investigate new uses of FDA-approved drugs or drug combinations." In Aim 1, we will systematically screen and prioritize drug repurposing candidates for AD/ADRD via both virtual transcriptome and Mendelian randomization, validating the candidates in multiple large clinical datasets. In Aim 2, we will identify, assess, and validate candidates from the literature. In Aim 3, we will evaluate and validate combinations of drug repurposing candidates. For all validations, we will assess the drug/drug pair’s benefits for prevention (i.e., pre-morbidity exposure) and treatment (i.e., post-morbidity exposure), including analyses stratified by dose, gender, and race. In Aim 4, we will make our findings, including deployable tools and codes, broadly available on the AD Knowledge Portal and related websites to facilitate future research.