MIAMI-. Although the survival rates of people diagnosed with some type of cancer have increased significantly in recent decades thanks to advances in technology. Doctors still face the great challenge of knowing when the tumor might resist chemotherapy treatment.
Now a group of scientists from the University of California San Diego School of Medicine has developed an Artificial Intelligence (AI) machine learning algorithm model which can facilitate this prediction.
The researchers analyzed a set of 718 genes commonly used in clinical genetic testing for cancer classification, using mutations within these genes for initial studies in tumor cells with their machine learning model.
Model test
They trained their model with publicly available drug response data, managing to identify 41 molecular sets, which are a group of collaborator proteins, where genetic alterations influence drug response.
They then tested it specifically in cervical cancer, in which about 35% of tumors persist after treatment. And they were able to successfully predict responses to cisplatin, one of the most common chemotherapy drugs.
AI Accuracy
According to the article, published in the magazine Cancer Discovery, The new algorithm was able to accurately identify tumors that were susceptible to the therapy, which was associated with better patient outcomes.
Besides, effectively recognized tumors that were likely to resist treatment and much of the underlying molecular machinery that drives treatment resistance.
The researchers highlighted that all cells, including cancer cells, rely on complex molecular machinery to replicate DNA as part of normal cell division. Most chemotherapies work by disrupting this DNA replication machinery in rapidly dividing tumor cells.
Although scientists recognize that the genetic makeup of a tumor greatly influences its specific response to drugs, “The multitude of mutations found within tumors has made the prediction of drug resistance more than challenging.“, they highlight.
They assure that This new algorithm overcomes this barrier by exploring how numerous genetic mutations collectively influence a tumor’s reaction to drugs that prevent DNA replication..
Trey Idekerlead author of the study, notes in a statement that doctors were previously aware of some individual mutations that are associated with resistance to treatment, “but these isolated mutations tended to lack significant predictive value.”
“This is because a much larger number of mutations can shape a tumor’s treatment response than previously thought. “Artificial intelligence closes that gap in our understanding, allowing us to analyze a complex range of thousands of mutations at once,” he explains.
The Ideker team specifies that Beyond predicting treatment responses, the model helped shed light on their decision-making process by identifying the sets of proteins that drive treatment resistance in cervical cancer..
They emphasize that the ability to interpret its reasoning is key to the success of the model and also to building reliable AI systems. “Unraveling the decision-making process of an AI model is crucial, sometimes as important as prediction,” he explained. Ideker.
They assure that The transparency of the model is one of its strong points“firstly, because it generates confidence, and secondly, because each of these molecular sets that have been identified becomes a new potential target for chemotherapy.”
They hope that this Artificial Intelligence will have broad applications, not only to improve current cancer treatment, but also to pioneer other new models.
@Lydr05
Source: With information from Cancer Discovery Magazine and EuropaPress
