IIT Madras researchers have developed a revolutionary method that can identify a person’s cancer-causing genes. Based on patient DNA profiles, the “Pivot,” an AI-based tool, will help clinicians create individualised treatment programmes for patients. The tool is built on a machine learning model that classifies genes into cancer-preventing tumour suppressors, oncogenes that cause cancer, and neutral genes.
- In order to identify the gene that causes cancer in an individual, or more specifically to identify unusual genes that cause cancer, the model will analyse data such as the frequency with which a gene mutates, its properties, the sorts of mutations, and the quantity of harmful mutations.
- It is well recognised that the patient’s overall health will suffer as a result of current cancer treatments.
- According to the research team, understanding the genes responsible for the development and spread of cancer in patients can aid in choosing the drug and therapy regimen that will have the greatest positive impact on their prognosis.
- The traditional method of tracing a cancer-causing gene involves analysing samples from several patients to determine how frequently the gene has mutated.
- If openly available cancer genome data from sites like ICGS had been of great assistance to the IIT-M scientists, then more Indian cancer genome data—which is still lacking on an exclusive local platform—is needed for the tool to be used effectively.
The International Cancer Genome Consortium (ICGC) and The Cancer Genome Atlas Program provide the majority of the genetic information used to feed the model (TCGA).