A new AI tool that predicts when cancer will become resistant to chemotherapy may lead to the development of ground-breaking new treatments, according to a study released today.
A machine learning system was used by researchers at the University of California San Diego School of Medicine to examine how tumors respond to medications.
The team effectively determined when tumors would respond to therapies by using its model to study cervical cancer, where about 35% of tumors do not respond to treatment. It is a significant discovery since most chemotherapies interfere with the intricate processes involved in DNA replication, which is a necessary component of regular cell division.
According to UC San Diego School of Medicine professor Trey Ideker, PhD, “hundreds of proteins work together in complex arrangements to replicate DNA.” “The way the entire tumor reacts to chemotherapy can be altered by mutations in any one component of this system.”
The model not only predicts treatment responses, but it also enables scientists to comprehend the decision-making process that goes into creating the predictions.
Ideker continues, “Understanding an AI model’s decision-making process is crucial, sometimes even more important than the prediction itself.” “We are hopeful that our model will find wide applications in both improving and developing new cancer treatments.”
