Endophenotype Network-based Approaches to Prediction and Population-based Validation of in Silico Drug Repurposing for Alzheimer's Disease

There have been more than 550,000 confirmed cases and over 22,000 deaths for COVID-19, the disease causedby the virus SARS-CoV-2, in the United States. Older individuals have the declined immune systems and ahigher mortality from COVID-19; furthermore, there are currently no effective antiviral medications againstCOVID-19. Drug repurposing, represented as an effective drug discovery strategy from existing drugs, offersemerging prevention and treatment strategies for COVID-19. SARS-CoV-2 requires host cellular factors forsuccessful replication during infection. Targeting virus-host protein-protein interactions (PPIs) offers an effectiveway for the development of drug repurposing (i.e., hydroxychloroquine (HCQ), melatonin, and indomethacin)for COVID-19 as demonstrated in our recent study (Cell Discovery 2020). Supported by the NIA R01, our teamare developing and implementing innovative network medicine and systems biology methodologies for drugrepurposing and drug combinations. We showed that HCQ was associated with a decreased risk of coronaryartery disease by reducing the expression of VCAM1 and IL-1β in human aortic endothelial cells. Exogenousmelatonin administration may be of particular benefit to COVID-19 older patients given an aging-related reductionof endogenous melatonin levels. Therefore, the central unifying hypothesis of this project is that an integrative,network medicine methodology that quantifies the interplay between the virus-host interactome and drugtargets in the human interactome network will offer highly repurposable drugs and clinically relevant combinationregimens for effective treatment of COVID-19. Aim 1 will test disease module hypothesis for prediction andvalidation of repurposable drugs for effective treatment of older individuals with COVID-19. We will utilize anetwork-based knowledge graph approach that incorporates not only virus-host interactions from SARS-CoV-2,but also public drug-target databases, the human protein-protein interactome, along with 24 millions ofpublications from PubMed database. Aim 2 will test the hypothesis that combining anti-inflammatory and antiviraltherapeutics for effective treatment of the underlying pulmonary and cardiovascular conditions in older individualswith COVID-19. We will utilize state-of-the-art pharmacoepidemiologic analyses to validate the clinical efficiencyof drug combinations (i.e., melatonin plus HCQ) in reducing incidence of pulmonary and cardiovascularconditions (including acute respiratory distress syndrome, pneumonia and lung injury) in older individuals, usinglarge-scale longitudinal Claims-Electronic Medical Record (EHR) patient databases, along with in vitroobservations in human aortic endothelial cell and pulmonary arterial endothelial cell models. To reduce theconfounding factors from patient databases, we will perform time-to-event pharmacoepidemiologic outcomeanalyses using large-scale de-identified patient EHRs from the Cleveland Clinic COVID-19 registry database.The successful completion of this project will offer clinically relevant repurposable drugs and combinationregimens for COVID-19 patients with aging-related pulmonary and cardiovascular conditions.1