Cancer treatment might just have received a big boost from an unlikely place. A machine learning algorithm is helping researchers at the University of Michigan and Indiana University exploit a weakness in cancer cells that allows them to grow uncontrollably.
Using lab mice, the researchers have identified backup genes only used by tumour cells. Using complex algorithms to track these backup genes, they can now target ovarian cancer with remarkable precision never before possible in medicine. The ML algorithm, they say, is easily adaptable to other forms of cancer as well. This is because the cellular behaviour of cancer cells that reveals these vulnerabilities is common in most cancers. So thanks to machine learning, doctors in the not-too-distant future can provide superior treatment plans for a variety of cancers.
An impressive breakthrough in precision medicine, the approach leverages collateral lethality known to occur in cancer cells to design drugs that only target cancer cells and spare normal ones. This has been the biggest pain point with the current regime of drugs used in cancer treatment. As tumour suppressor genes are deleted, a correspondingly large number of genes without direct roles in tumour progression are also deleted.
To avoid death due to these deletions, the cancer cells find a paralog. The latter helps keep the cancer cell alive by performing similar functions to the deleted genes. Using a mathematical model of cancer cells’ metabolism, scientists can target and shut down the right paralogous metabolic pathways. The algorithm correctly sorted through multiple options and successfully predicted how a cell missing its paralog would turn to its backup gene.
Meanwhile, researchers at Indiana University School of Medicine helped validate the findings in the lab. They developed genetically modified cell and animal models of ovarian cancers with deletions. Six out of six mice tested showed complete cancer remission! It is likely possible machine learning is affording our generation the rare chance to witness the beginning of the end of cancer.