Does artificial intelligence need companionship to assist in drug discovery? The Kirsten rat sarcoma virus study

Abstract

Abstract In this Opinion article, we confront the role of artificial intelligence (AI) in targeting and understanding resistance to targeted therapy using the most frequently mutated oncoprotein family in human cancer, rat sarcoma virus guanosine triphosphate hydrolases (RAS GTPases), here Kirsten RAS (KRAS), as an example. Aberrant regulation of the active GTP-bound state of KRAS is associated with tumourigenesis, aggressive disease, and poor prognosis. KRAS mutations (eg, G12C, G12D, G12V, G13D, inter al.) are drivers of numerous cancer types, including non-small cell lung, colorectal, and pancreatic cancers. These mutations have shown to play a significant role in cell behaviour and response to treatment. Since its discovery in the 1980s, it has been recognized that over-expression of KRAS and other RAS family members induces resistance to radiotherapy. Moreover, over the years preclinical and clinical studies showed that tumours with KRAS mutations exhibit different treatment sensitivities compared to tumours with wild-type KRAS.