Scientists Develop AI “Fingerprint” Technology to Revolutionize Cancer Drug Development

Researchers at The Institute of Cancer Research in London have unveiled a groundbreaking AI ‘fingerprint’ technology that accurately reveals how cancer cells respond to new drugs through an observation of changes in their shape.

This innovative approach can potentially cut years from the drug development process and has been published in Cell Systems.

Enhancing the Drug Development Process

The new technology allows for rapid assessment of new drugs’ effectiveness at targeting specific cancer types, facilitating faster patient access to effective treatments. The AI’s ability to predict drug responses by analyzing 3D cell images enables researchers to design more effective clinical trials tailored to sub-types of cancer, ultimately minimizing costly failures and saving millions in research costs.

Detailing the 3D Imaging Structure

The AI was trained using nearly 100,000 3D images of melanoma cells, taken through leading-edge microscopy techniques. In contrast to previous methods that relied on 2D imaging, this technology incorporates real cell shapes, providing a more accurate representation that the body typically presents.

With this new tool, researchers can predict drug interactions with up to 99.3% accuracy, even distinguishing between drugs that target different proteins yet produce similar effects on cells.

Speeding Up the Drug Development Timeline

Traditionally, drug development spans over a decade. However, the researchers estimate that leveraging AI technology early in the process could compress a three-year preclinical phase into just three months, expediting trials by as much as six years.

The researchers have indicated that their AI tool is versatile and can be applied to other cell types, including stem cells and red blood cells, enhancing its utility across various diseases.

Implementing AI in Drug Discovery at ICR

The Institute of Cancer Research plans to integrate this AI technology into its drug discovery initiatives, aiming to develop new protein-targeting drugs. Alongside this, the researchers have set up a spinout company named Sentinal4D to advance this technology further.

Professor Chris Bakal emphasized the utility of the AI technology, stating, ‘3D cell shape is like a fingerprint of cellular state and function—it’s an untapped reservoir of information. Using AI, we can decode this fingerprint.’

He added, ‘The tool is powerful enough to streamline the drug discovery process, ultimately benefiting patients who need new treatment options rapidly.’

Future Directions in Cancer Treatment

With ongoing research, the team is set to explore biomarkers discovered through the AI’s analysis, aiming to develop the next generation of cancer drugs aimed at saving lives.