Scientists have developed a revolutionary AI "fingerprint" technology that can accurately show how cancer cells respond to new drugs, by simply observing changes to their shape.
The new technology, which has been developed by a team at The Institute of Cancer Research, London, will allow researchers to quickly assess the ability of new drugs to reach their intended target, slashing years off the drug development process -- allowing new drugs to reach patients faster.
The scientists believe their approach could also save millions of pounds by reducing investment and effort in projects that go on to fail. The research is published in the journal Cell Systems.
Using 3D imaging of cells
Crucially, the tech helps scientists match the right drugs to the right patients, by enabling them to design clinical trials for specific cancer sub-types at a much earlier stage -- thereby avoiding costly clinical trial failures.
The team from The Institute of Cancer Research (ICR), trained the AI technology using almost 100,000 3D images of melanoma skin cancer cells -- taken with cutting-edge microscopy -- and geometric deep learning to analyze the information about the shape of the cells.
Previous technologies have only been trained on flat, 2D images of cells on a microscope slide -- which don't take into account the 3D shape of a cell, as it appears in the body.
The team treated the cells with a variety of drugs and used their newly created AI tool to learn which shape changes were caused by each drug.
They showed that the tool could predict which drug was being used on the cells with up to 99.3% accuracy and it could even distinguish between the shape changes caused by drugs which, although they target different proteins, ultimately have very similar effects on the cell.
The researchers showed that the AI technology was accurately learning the underlying biochemical changes occurring when melanoma cells were treated with certain drugs. It was able to identify important proteins that the team are now exploring as potential targets to develop new drugs.
Cutting out stages in the drug development process
The team also showed their AI tool worked for other cell types -- including red blood cells, cells in brain vessels, and stem cells -- indicating that other diseases could benefit from this technology.
Developing a new drug usually takes 10 to 12 years. However, the ICR team believe that using their AI technology early in this process could cut out numerous steps in the preclinical phase -- slashing it from three years to three months -- and reduce the time taken to trial new drugs by up to six years, as patients most likely to benefit could be determined earlier, and side effects could be predicted.
The AI tool outperformed other similar algorithms as it is the first to use 3D information on a cell's shape -- the full picture of the cell, as it would appear in a body -- instead of only 2D information on a microscope slide. The tool was also trained to take into account the variability in a population of cells, while other algorithms look either at single cells or take an average of cell shape across the population.
Implementing the tool into the ICR's drug discovery research
The researchers will work with teams within the ICR's Center for Cancer Drug Discovery to implement their AI technology into the process of discovering targeted protein degraders -- a new type of drug that co-opts a cell's natural waste disposal system to remove the offending protein.
The research team has also patented their tool and set up a spinout company, Sentinal4D, to take the innovation into the next phase and implement it into the drug discovery and development pathway.
Sentinal4D is the latest spinout company to be announced by the ICR and follows recent spinout successes including the foundation of Monte Rosa Therapeutics, which is now listed on New York's NASDAQ stock exchange.
'The tool is so powerful that we will be able to streamline the drug discovery process'
Professor Chris Bakal, Professor of Cancer Morphodynamics at The Institute of Cancer Research, London, said, "3D cell shape is like a fingerprint of cellular state and function -- it's a previously untapped reservoir of information. Using AI, we can decode this fingerprint and reveal how cells respond to drugs.
"The tool that we've created is so powerful that we will be able to streamline the years-long drug discovery process, saving both time and money. Patients with cancer need new treatment options as quickly as possible, so speeding up this process will be hugely valuable."
Dr. Matt De Vries, co-founder and chief technology officer of Sentinal4D, said, "With the AI tool we've created, it will be possible to predict how effective a drug will be and if there are likely to be any side effects. The tool could work for a range of diseases, as we've shown that it will pick up the changes in shape for a number of different cell types and drugs.
"Our new company, Sentinal4D, aims to use this tool to eliminate the guesswork and increase the chances of success in the subsequent phases of drug development -- to bring treatments to patients sooner."
Professor Kristian Helin, chief executive of The Institute of Cancer Research, London, said, "The ICR is dedicated to discovering new drugs to meet the challenges of cancer evolution and drug resistance, so cancer patients have more treatment options -- extending and saving lives.
"This latest technology builds on years of work at the ICR to understand cancer cell shape and to use artificial intelligence to analyze data. I look forward to seeing this technology being used to develop new medicines that have a real impact for people with cancer."
