A team from The Institute of Cancer Research have used artificial intelligence to predict how cancers will progress and evolve.
The new technique, called REVOLVER (Repeated evolution of cancer), picks out patterns in DNA mutations within cancers and uses the information to forecast future genetic changes.
Study leader Dr Andrea Sottoriva, Team Leader in Evolutionary Genomics and Modelling at The ICR said: “We’ve developed a powerful artificial intelligence tool which can make predictions about the future steps in the evolution of tumours based on certain patterns of mutation that have so far remained hidden within complex data sets.
“With this tool we hope to remove one of cancer’s trump cards – the fact that it evolves unpredictably, without us knowing what is going to happen next. By giving us a peek into the future, we could potentially use this AI tool to intervene at an earlier stage, predicting cancer’s next move.”
If clinicians can predict how a tumour will evolve, they can intervene earlier to stop cancer in its tracks before it has had a chance to evolve or develop resistance, increasing the patient’s chances of survival. The team found a link between certain sequences of repeated tumour mutations and survival outcome, suggesting repeating patterns of DNA mutations could be used as an indicator of prognosis.
The new machine learning technique transfers knowledge about tumours across similar patients. This method identifies patterns in the order that genetic mutations occur in tumours that are repeated both within and between patients’ tumours, applying one tumour’s pattern of mutations to predict another’s.
The researchers used 768 tumour samples from 178 patients reported in previous studies for lung, breast, kidney and bowel cancer, and analysed the data within each cancer type respectively to accurately detect and compare changes in each tumour.
By identifying repeating patterns and combining this with current knowledge of cancer biology and evolution, the scientists could predict the future trajectory of tumour development. The work is published in the journal Nature Methods.