How Genomics, Epigenetics and AI Are Changing Cancer Detection and Treatment
Cancer is no longer seen as a single genetic error but as a complex, multi-layered disease shaped by DNA mutations, epigenetic changes and even patterns in medical images. New research at the Centre for Computational Natural Sciences and Bioinformatics (CCNSB) at IIIT Hyderabad is bringing these layers together to move closer to early detection and truly personalised cancer care.
A century ago, scientists believed cancer began with a single mistake in a cell. In 1914, the somatic mutation theory proposed that abnormalities in a cell’s DNA could trigger uncontrolled growth. Over time, this idea expanded. Researchers discovered oncogenes that drive cancer and tumour suppressor genes that normally prevent it. Later theories showed that cancer does not arise from rogue cells alone – the surrounding tissue environment, viruses, carcinogens, and cellular stress also play critical roles.
“People have been talking about the origin of tumours since the early 1900s, but over time we realised that cancer cannot be explained by mutations alone. Today, cancer is understood as a multifactorial disease, shaped by genetics, gene regulation, environment and time,” observes Prof. Nita Parekh, Professor of Bioinformatics, IIIT-H.
Reading the Genome: What DNA Variations Tell Us
At the heart of modern cancer research is genomics – the study of variations in DNA. “One part of my work looks at the role of genetic variations in tumorigenesis,” remarks Prof. Parekh referring to some small changes such as a single-letter swap in the genetic code. Or sometimes large changes, involving missing segments, duplications, inversions or even the fusion of genes from different chromosomes. These variations can silently accumulate or dramatically alter how cells behave. “By analysing cancer genomes in detail, we can identify which mutations matter, which pathways they disrupt, and how different cancers – or even subtypes of the same cancer – behave very differently,” states Prof. Parekh.
Studying Aggressive Blood Cancers
One focus of Prof. Parekh’s research has been on Diffuse Large B-Cell Lymphoma (DLBCL), an aggressive blood cancer. DLBCL has two main subtypes, each with very different outcomes. “When we looked at subtype-specific genetic variations, we could clearly see why some patients respond well to treatment while others do not,” she says. By profiling genetic variations across cancer cell lines, Prof. Parekh’s team identified subtype-specific mutations, disrupted pathways and novel biomarkers for prognosis and therapy. These insights enable genome-guided treatment strategies, where therapies are chosen based on the patient-specific mutational profile.
Beyond DNA: The Epigenetic Layer
“Genes alone do not tell the full story,” says Prof. Parekh explaining how cells also use epigenetic mechanisms – chemical switches that turn genes on or off without changing the DNA sequence. “Epigenetic regulation is the ongoing part of my work,” she mentions. One of the most important of these is DNA methylation. In cancer, tumour-suppressing genes can be silenced, or cancer-promoting genes can be activated, simply by altering these chemical tags. Prof. Nita Parekh and her team are now studying the methylation status of not just gene promoters, but enhancers, gene bodies and non-coding RNAs as well – uncovering regulatory changes that often appear before a tumour fully develops, making them powerful candidates for early detection.
The Hidden Players: Non-Coding RNAs
Another emerging layer involves non-coding RNAs, including microRNAs and long non-coding RNAs (lncRNAs). “These RNAs do not code for proteins, but they play a major role in regulating gene expression,” remarks Prof. Parekh. MicroRNAs and long non-coding RNAs (lncRNAs) act as regulators, fine-tuning gene expression. In cancer, these molecules can suppress protective genes or activate harmful ones. Mapping how they interact – forming complex regulatory networks – would offer new insights into why cancers behave differently across patients and subtypes.
Integrating Data for Early Breast Cancer Detection
One other type of cancer that the IIIT-H team is actively looking at is breast cancer. It is one of the most common cancers worldwide. “Breast cancer is not one disease; it has several subtypes, and identifying them early makes a critical difference,” says Prof. Parekh. By combining DNA methylation data, RNA expression profiles and machine learning, the researchers have identified molecular signatures that can distinguish cancer subtypes, predict patient risk and survival and also point to early diagnostic markers. These findings pave the way for liquid biopsies – simple blood tests that detect cancer signals long before symptoms appear.
Seeing Cancer Through Images
The research also extends beyond molecular data into medical imaging. “I’m also working on mammography data analysis, which we are developing in parallel,” notes Prof. Parekh. Her team has developed large, curated datasets of mammograms to train AI models that can detect abnormalities early, segment suspicious regions, classify tumours as benign or malignant and generate preliminary clinical reports. This work aims to support radiologists, reduce diagnostic delays, and improve screening accuracy – especially in resource-constrained settings.
Why This Research Matters
India faces unique cancer challenges – genetic diversity, younger age of onset, and delayed diagnosis. Research rooted in Indian data is essential for building accurate, equitable solutions. “Understanding cancer requires looking at genetics, epigenetics and gene expression data together – not in isolation,” reiterates Prof. Parekh. By integrating genomics, epigenetics and AI-driven imaging, this research is moving the future of cancer care closer to precision medicine – where treatment is guided by the unique biological profile of each patient – to deliver better outcomes for all
