RepurNet: A network medicine framework for systematic drug repurposing and disease vulnerability discovery

Drug repurposing has become a versatile strategy in rapidly identifying drug candidates to treat disease. Using the early COVID-19 pandemic as an example, we perform a systems-scale in silico screen to identify compounds that induce a gene regulatory signature reverse of those found in the disease state. We design a disease module integrating protein interactions to suggest rational combinations of drug candidates to counteract functional changes during disease. We present this strategy as RepurNet, a high-throughput in silico framework to rapidly prioritize compounds in early drug discovery. We determine network similarities between distinct diseases that could share common morbidities to illustrate the extension and application of this approach to other diseases, such as cancer. Together, we present a generalizable integrative systems biology and network medicine approach to unveil signaling perturbation and specificity in disease, providing optimized repurposed drug candidates, with broad implications for precision medicine in various disease contexts.