Optimize cancer treatment with new AI-powered protein mapping tool

Scientists have developed a new AI tool to map the function of proteins within cancerous tumors. This allows clinicians to more accurately determine treatment goals.

Cancers such as clear cell renal cell carcinoma (ccRCC) respond differently to existing treatments, making it difficult to identify the appropriate drug treatment plan for each patient.

For example, the cancer drug Belzutifan was recently approved for the treatment of ccRCC, but the response rate in patients with the most common form of this condition is only 49%.

To better understand why some patients respond better than others, researchers at the University of Bath and the University of Nottingham investigated hypoxia, a key target in ccRCC that is inhibited by belzutifan. We investigated the function of inducible factor alpha (HIF2α).

Previous studies have shown that HIF2α levels do not necessarily correspond to tumor grade, and counterintuitively that HIF2α activity decreases when higher levels of the protein are present.

This means that administering high doses of belzutifan may expose patients to expensive and toxic treatments that may not be effective, and may even increase drug resistance in tumors. It means that.

An interdisciplinary team of biophysicists, biologists, and computational scientists has devised a new tool called FuncOmap that maps the functional state of targeted tumor proteins onto tumor images.

This will allow clinicians to directly visualize where tumor proteins are interacting within the tumor, enabling more accurate diagnosis and providing the optimal treatment for each patient. .

Professor Banafsheh Larijani, director of the Center for Therapeutic Innovation at the University of Bath, co-led the study. She said: “Responses to drugs vary greatly from person to person. Therefore, we can individually predict how patients will respond to drugs so that they can be administered the lowest possible dose to minimize side effects while still achieving efficacy.” It is important to be able to adjust your treatment accordingly.

“By precisely mapping the functional state of tumor proteins in patient tumor sections, our new computational analysis tool can help clinicians improve patient stratification and enable personalized medicine. .”

The team is currently collaborating with Dr. Amanda Killern's lab at Stanford University School of Medicine, as well as other surgeons and clinicians, to develop and optimize the tool for clinical use.

Professor Eamonn O’Neill, Head of the School of Computer Science at the University of Bath and Director of the UKRI Doctoral Training Center in Responsible, Accountable and Transparent AI (ART-AI), said: The essence of working across disciplines.

“This will be under the umbrella of the UKRI Center for Accountable and Transparent PhD Training in Artificial Intelligence, which will integrate computer science, biology and physics to improve clinical decision-making and personalized research in cancer. “We provide image analysis that has the ability to directly inform clinical outcomes. Like any other disease, treatment is also required.”

The excitement of this paper lies not only in the research content reported, but also in the way it combines the research fields of biophysics and translational medicine with modern computational science, transforming the research into a valuable tool for clinical settings. This is where it explains how it accelerates. This further increases the value gained from interdisciplinary research. ”

Jonathan Knight FRS Professor, Vice President (Enterprise).University of Bath

This research was funded by the UKRI Center for Doctoral Training in Accountability and Transparent Artificial Intelligence (ART-AI). [grant number EP/S023437/, Medical Research Council and University of Bath Alumni Fund, and is published in the journal BJC Reports.