AI-Based OncoMark Framework Brings Fresh Insight Into How Cancer Cells Grow and Spread

A team of Indian scientists has developed an artificial intelligence–powered framework that could transform how doctors understand and treat cancer. The tool, called OncoMark, goes beyond examining a tumour

from the outside. It studies the hidden internal mechanisms that cancer cells use to grow, survive, and spread, allowing researchers to analyse cancer from within.The findings were published in Communications Biology in October, marking a significant step toward more precise and personalised cancer therapy.

Why Understanding Internal Cancer Processes Matters

Cancer diagnosis today focuses heavily on what is visible: the size of a tumour, where it is located, and whether it has spread. While these features are important, they do not explain what is happening at the molecular level that causes a cell to behave abnormally.Most clinical evaluations skip over the hallmarks of cancer—those core traits scientists pinned down back in 2011. These hallmarks mark out the key biological shifts that turn a regular cell into a cancer cell. We’re talking about things like unchecked cell growth, dodging cell death, invading nearby tissues, and slipping past the immune system. At first, researchers counted six of these hallmarks. Now the list stands at ten.These internal behaviours define cancer at its core. For example, a healthy cell has a built-in mechanism that determines when it should stop functioning and die. A cancer cell disables this mechanism, allowing it to continue living. It also switches off growth-control systems, enabling rapid and excessive multiplication. Such molecular changes begin long before a tumour becomes visible in scans.

How OncoMark Works

OncoMark uses artificial intelligence to analyse these hallmark processes at a genetic level. The model was trained using genetic information from 3.1 million cancer cells across 14 types of cancer. This allowed it to learn how internal biological processes work together to fuel tumour development and resistance to treatment.The researchers tested the system on five separate datasets, and the AI consistently delivered accuracy levels of at least 96 percent in predicting how aggressively a cancer cell could grow, spread, or resist treatment. This enables the tool to create a detailed molecular profile of a tumour by estimating the activity of each hallmark.

Toward More Personalised Cancer Care

OncoMark digs into the molecular workings inside tumors, letting doctors see just how aggressive a cancer really is. Instead of just looking at what’s on the surface, oncologists start building treatment plans around the actual biology driving the disease for each patient.This shift opens the door to more personalised, targeted therapies that improve outcomes and spare patients from unnecessary treatments.

What Comes Next

The research team aims to integrate OncoMark into clinical practice so doctors can use the tool directly for patient evaluation. They also plan to expand their work to blood cancers and rare cancer types, which behave differently from solid tumours and may require specialised analysis.Although still in development, OncoMark represents a major step toward a future where cancer treatment considers not just what doctors can observe externally, but also the molecular strategies a tumour uses to survive. With further refinement and clinical adoption, it may help pave the way for a new generation of highly personalised cancer therapies.