Indian Researchers Develop GenAI Platform to Trace Potential Root Cause of Cancer
New AI Platform May Help Trace Cancer-Causing Exposures
Researchers from Delhi's Indraprastha Institute of Information Technology have developed a generative AI platform that could help researchers understand the potential causes behind cancer-related DNA damage.
Called MutAIverse, the platform analyses molecular information from tumour samples and attempts to trace damaged DNA back to possible environmental chemicals or toxins that may have contributed to the damage.
The researchers describe it as the first generative AI platform of its kind designed to expand the search for potential cancer-related DNA adducts.
Platform Tested on Head and Neck Cancer Patients
MutAIverse was experimentally tested using biopsy samples from 27 head and neck cancer patients in Guwahati, Assam.
The patients had a history of consuming smokeless tobacco. The samples were collected in collaboration with the Dr Bhubaneswar Borooah Cancer Institute, while DNA adductomics was conducted at the National Institute of Pharmaceutical Education and Research.
The research team also included scientists from NIPER and the CSIR-Institute of Genomics and Integrative Biology.
How MutAIverse Works
The platform uses a machine-learning-based backtracking tool called AdductLinker.
When DNA damage is detected in a tumour sample, AdductLinker attempts to work backwards and identify the environmental chemical or toxin that may have produced the damage.
Researchers compare the process to tracing clues left behind at a scene.
The platform uses mass spectrometry data from a patient's tumour biopsy as an input and generates information about potential exposure as an output.
AI Expands DNA Adduct Reference Library
One of the significant features of MutAIverse is its ability to expand reference libraries of DNA adducts.
DNA adducts are chemical compounds that can attach to DNA and alter its structure. Some forms of DNA damage may contribute to mutations and genomic instability.
According to the researchers, the platform helped expand an existing library from fewer than 400 reference DNA adducts to more than three lakh potential structures.
The research has been published in the Journal of Cheminformatics.
Does the Technology Prove What Caused Cancer?
Experts have cautioned that the platform does not currently prove that a particular exposure caused a patient's cancer.
Detecting a DNA adduct only shows that a chemical has interacted with DNA. It does not establish that the damage caused a mutation or that the mutation eventually resulted in cancer.
Some DNA damage can be repaired by cells, while other forms may persist and contribute to genomic instability.
More Clinical Evidence Needed
Experts say additional experimental, clinical and epidemiological evidence will be required before the technology can establish a direct connection between a specific chemical exposure, DNA damage and cancer.
Kedar Natarajan, a computational biologist and professor at Ashoka University, noted that understanding cancer-causing exposures could be important for studying cancer severity, mutation patterns, disease management and potential treatment strategies.
However, the platform in its current form should not be considered a clinical diagnostic tool for hospitals or community use.
Potential Role in Cancer Prevention
Researchers believe that identifying possible exposure to tobacco products, occupational chemicals or other environmental toxins could eventually help patients avoid further exposure.
It could also help alert people living or working in similar risk environments.
Overall, MutAIverse represents a complementary research tool rather than a definitive cancer diagnostic system. With further validation, researchers believe it could be used alongside genome sequencing and other methods to build a more detailed understanding of how individual cancers develop.
