A research group from IIT Madras developed an AI-based tool that can predict cancer-causing genes in a person. Doctors can devise personalised treatment strategies using the new tech.
Indian Institute of Technology, Madras, research group has named this tool PIVOT. They have designed it to foretell the presence of cancer-causing genes from a sample. They use it to operate information on mutations, gene expression, gene variations, and biological network perturbations.
Dr Kartik Raman, Core Member, RBCDSAI, IIT Madras, stated that doctors should not treat a complex disease like cancer with one treatment. Cancer treatments include personal medications. The AI tool can help in making distinctions between patients.
Cancer: Leading Cause Of Death
Moreover, PIVOT, based on a machine learning model, can classify genes such as tumour, suppressor genes, oncogenes or neutral genes. According to WHO, the worldwide leading cause of death is cancer. In 2020, one in six deaths were due to cancer, despite the COVID-19 pandemic.
The available cancer treatments, including chemotherapy and radiation therapy, are allegedly detrimental to a patient’s health. Thus, the information on genes responsible for the initiation and progression of cancer helps to identify suitable drug combinations.
The new AI-based tool predicted the existing oncogenes and tumour-suppressor genes like TP53, and PIK3CA, among others. It also identified new cancer-related genes like PRKCA, SOX9, and PSMD4.
However, the makers of the AI-based model say that the tool meant for different types of cancer is at a nascent stage. It detects breast, invasive carcinoma, colon adenocarcinoma, and lung adenocarcinoma. But, PIVOT helps to push boundaries and presents chances for empirical research based on the genes, according to Malvika Sudhakar, Reserach Scholar from IIT Madras.
Furthermore, the team plans to extend PIVOT to many cancer types. The team is also working on a list of personalised cancer-causing genes that can help to identify the suitable drug for patients based on their cancer profile.