Ovanessians, Armand2023-11-192023-11-19https://hdl.handle.net/11124/178547https://doi.org/10.25676/11124/178547In the absence of effective antiviral drugs for the treatment of COVID-19, we designed a novel k-partite graph learning method for data matrix completion to uncover hidden viral-host interactions and drug-target interactions. Our framework produces a holistic map of the pathogenesis of COVID-19 and unveils novel FDA approved drug candidates for COVID-19.articlesengCreative Commons CC-BY License or the Creative Commons CC-BY-NC License.Text