According to the World Health Organisation, cancer accounted for nearly one in six deaths in 2020. Cancer can occur due to mutations in oncogenes or tumour suppressor genes or both. However, not all mutations result in cancer. Therefore, it is important to identify the genes causing cancer to devise personalised treatment strategies.
IIT Madras researchers have developed an artificial intelligence-based tool, ‘PIVOT’, that can predict cancer-causing genes in an individual. The prediction is based on a model that utilises information on mutations, expression of genes, and copy number variation in genes and perturbations in the biological network due to an altered gene expression.
The research was led by Raghunathan Rengaswamy, Dean (Global Engagement), and professor, department of chemical engineering, IIT Madras; Karthik Raman, associate professor, Bhupat and Jyoti Mehta School of Biosciences, and a core member, Robert Bosch Centre for Data Science and Artificial Intelligence (RBCDSAI), IIT Madras; and Malvika Sudhakar, a research scholar at IIT Madras.
Raman said, “Cancer, being a complex disease, cannot be dealt with in a one-treatment-fits-all fashion. As cancer treatment increasingly shifts towards personalised care, such models that help pinpoint differences between patients can be very useful.”
Current cancer treatments are known to be detrimental to the overall health of the patient. Knowledge of the genes responsible for the initiation and progression of cancer can help determine the combination of drugs and therapy most suitable for a patient’s recovery.
Although there are tools available to identify personalised cancer genes, they use unsupervised learning and make predictions based on the presence and absence of mutations in cancer-related genes. This study is the first one to use supervised learning and takes into account the functional impact of mutations while making predictions.
The researchers have built AI prediction models for three types of cancer: Breast Invasive Carcinoma, Colon Adenocarcinoma and Lung Adenocarcinoma. They plan to extend the tool’s use to many more cancer types. The team is also working on a list of personalised cancer-causing genes that can help identify the most suitable drugs for individual patients.