New Delhi: At one time it was believed that DNA mutations were the sole cause for cancer. It is now recognised that cancer is a multifactorial disease influenced by genetics, gene regulation, environmental factors and time. Genomics research focuses on DNA variations that drive tumorigenesis. These include small ‘single-letter’ changes in the genetic code as well as larger structural alterations such as missing segments, duplications, inversions and gene fusions from different chromosomes. Detailed analysis of cancer genomes have been used to identify key mutations, disrupted pathways and differences in behaviour across cancer types and subtypes. In aggressive blood cancers such as Diffuse Large B-Cell Lymphoma, subtype-specific genetic variations explain varying treatment responses.

Profiling these variations across cancer cell lines has revealed novel biomarkers for prognosis and therapy selection, enabling genome-guided strategies tailored for individual mutational profiles. Epigenetic mechanisms are another regulatory layer, that do not alter DNA sequences. DNA methylation acts as chemical switches that silence tumour-suppressing genes or activate cancer-promoting ones. Current studies examine methylation not just as gene promoters but also as enhancers, gene bodies and non-coding RNAs. These changes occur in the early stages of cancer, before full tumour development, positioning them as candidates for early detection. Non-coding RNAs, including microRNAs and long non-coding RNAs regulate gene expression by supressing protective genes or activating harmful ones. Mapping their interactions forms complex networks that explain patient subtype and differences in cancer behaviour.

Detecting cancer before symptoms emerge

For breast cancer, integration of DNA methylation data, RNA expression profiles and machine learning has been advanced by researchers at IIIT Hyderabad to identify molecular signatures distinguishing subtypes, predicting risk and survival, and highlighting early diagnostic markers. This supports liquid biopsies using blood tests to detect signals before symptoms emerge. The researchers have also applied AI to mammography imaging. Curated datasets are used to train models to detect abnormalities early, segment suspicious regions, classify tumours as benign or malignant and generate preliminary reports.

These tools aim to assist radiologists, reduce delays and improve accuracy in resource-limited settings. Integrating genomics, epigenetics, gene experession and AI-driven imaging advances precision medicine, where treatments match the unique biological profile of each patient, improving outcomes. The researchers have used Indian data, which addresses local challenges such as genetic diversity, younger onset of cancer and delayed diagnosis.