| The COVID-19 pandemic has caused millions of deathsand massive societal distress worldwide. Therapeutic solu-tions are urgently needed, butde novodrug developmentremains a lengthy process. One promising alternative iscomputational drug repurposing, which enables the pri-oritization of existing compounds through fastin silicoanalyses. Recent efforts based on molecular docking,machine learning, and network analysis have producedactionable predictions. Some predicted drugs, targetingviral proteins and pathological host pathways are under-going clinical trials. Here, we review this work, highlightdrugs with high predicted efficacy and classify theirmechanisms of action. We discuss the strengths andlimitations of the published methodologies and outlinepossible future directions. Finally, we curate a list ofCOVID-19 data portals and other repositories that could beused to accelerate future research |