MutAIverse Decodes DNA ‘Footprints’ to Uncover Possible Cancer-Linked Exposures

Indian researchers have developed MutAIverse, a generative AI platform designed to help trace environmental exposures that may be linked to cancer-related DNA damage. Developed by researchers at Indraprastha Institute of Information Technology, Delhi, with collaborators from NIPER and CSIR-IGIB, the platform shifts attention from simply diagnosing cancer to investigating what may have contributed to its development.

MutAIverse analyses mass-spectrometry data generated from tumour biopsies. The molecular fingerprints reveal chemical modifications in DNA, including DNA adducts —structures formed when reactive chemicals or their metabolites attach to DNA. The platform uses generative AI and mechanistic simulations to expand its reference library from fewer than 400 adducts to more than three lakh potential structures.

Its machine-learning tool, AdductLinker, works backwards from an observed DNA adduct to identify a possible parent genotoxin or carcinogen. It can also model how chemicals are metabolised inside cells and generate an exposure profile, including a suspected chemical, confidence score, damaged-DNA structure and intermediate metabolites.

Researchers tested the system on biopsy samples from 27 head-and-neck cancer patients in Guwahati with histories of smokeless-tobacco consumption. DNA-adductomics revealed selective enrichment of both known and novel DNA adducts, with a subset validated through MS/MS analysis. The findings provided evidence that the platform can connect DNA-damage signatures with plausible exposures associated with smokeless tobacco.

The potential applications are significant. MutAIverse could help researchers investigate cancer-causing exposures involving tobacco, pollution and occupational chemicals, identify biomarkers of exposure, improve understanding of cancer mechanisms and eventually support prevention by helping people reduce harmful exposures.

However, MutAIverse is not yet a clinical diagnostic tool. A DNA adduct proves that a chemical interacted with DNA, not that the damage became a permanent mutation or caused cancer. Cells can repair some damage, while other damage may contribute to genomic instability.

Before hospital use, researchers need larger clinical studies, laboratory confirmation of predicted adducts, prospective testing of accuracy and reproducibility, and long-term epidemiological studies. They must establish the chain from chemical exposure to DNA damage, mutation and cancer, while accounting for other biological and environmental factors.