DrugsiderAI (subproject 3 and 4): AI-based methods for synergistic exploration of disease symptoms and drug side effects

Abstract

Medications may cause unintended side effects that can affect the overall health of patients. The mechanism of drug adverse reaction might involve multiple pathways, where the drug could interact with any off-target protein within relevant pathways. Moreover, some drug side effects resemble symptoms of underlying diseases, revealing shared or overlapping biological pathways between the adverse drug responses and disease mechanisms. Thus, understanding the genetic factors prompting adverse drug responses is crucial to minimize the risk in human health. To this end, we build variant effect prediction models that predict the functional effects of multiple mutations, with the goal of better understanding their potential role in drug side effects, symptoms, diseases or related clinical outcomes.