Machine Learning Empowered Mechanism Prediction for Immunosuppressive Therapy of Sepsis

Description

Sepsis causes high morbidity and mortality, and therapy relies only on supportive treatment, including antibiotics and fluids. Currently, there are no FDA-approved specific drugs to treat sepsis. Aiming to develop effective sepsis treatments, we have identified a critical role of some immunosuppressive drugs that block host immune responses against invading microbes. However, whether these drugs share common functional mechanisms, and if yes, whether we can harness these mechanisms to improve sepsis therapy is unknown. This proposal will screen more potential anti-sepsis drugs. We will use the preliminary results to develop a meta-learning drugtarget bioactivity prediction framework to identify promising targets. The establishment of this framework will improve the identification and development of anti-sepsis drugs and, more importantly, renovate the platform for drug development in the future.