Indian researchers have developed MutAIverse, a generative AI platform designed to help identify environmental exposures that may have contributed to cancer by analysing DNA damage. The platform was tested on biopsy samples from 27 patients with head and neck cancer in Guwahati and identified patterns associated with smokeless tobacco exposure.
Linking DNA Damage to Environmental Exposure
The platform uses a machine-learning tool called AdductLinker to work backwards from damaged DNA and identify chemicals that may have interacted with it. It also expanded existing DNA adduct reference libraries from fewer than 400 entries to more than three lakh potential compounds.
As per ETHealthworld, the researchers believe the technology could help move cancer research beyond diagnosis by providing insights into possible environmental and occupational exposures. The system analyses mass spectrometry data from tumour biopsies and generates information about potential toxins, damaged DNA structures and intermediate chemicals produced in the body.
More Validation Needed
However, experts caution that identifying a DNA adduct does not prove that the associated chemical caused the cancer. Further experimental, clinical and epidemiological studies are needed to establish a direct causal link.
At present, MutAIverse remains a research tool rather than a clinical diagnostic system. With further validation, researchers suggest that it could complement genome sequencing and other approaches to improve understanding of how individual cancers develop.