AlphaGenome Decodes 98% of Human Genome, Google DeepMind’s Model Paving Way for Advanced Cancer and Rare Disease Treatments

Mumbai, February 5: Google DeepMind has announced the development of AlphaGenome, a breakthrough artificial intelligence model capable of decoding the "dark matter" of the human genome. Published in the journal Nature, the AI identifies how minute changes in DNA sequences influence the development of chronic diseases, including cancer and rare genetic disorders.

The model represents a significant evolution from the previous AlphaFold system, which primarily focused on protein structures. While AlphaFold could only interpret the 2% of the genome responsible for protein production, AlphaGenome analyzes the remaining 98%, often referred to as "genetic switches" that control when and how genes are activated. 'Prevent Cancer Before It Starts': New WHO Study Maps Risks.

AlphaGenome AI Technology

AlphaGenome is engineered to process DNA sequences of up to 1 million base pairs at once. This high-resolution capability allows the AI to predict 11 critical biological features simultaneously, such as gene expression levels and DNA structural folding. DeepMind researchers described DNA as the "recipe" for life, noting that the new model allows scientists to understand the complex grammar of genetic sequences.

The AI utilizes a technique called "knowledge distillation," which merges data from multiple models into a single, efficient system. During performance evaluations, AlphaGenome outperformed existing models in 25 out of 26 tests. It successfully identified approximately 5,900 human genetic signals, offering a more comprehensive map of the human body’s internal regulatory systems than previously possible.

Genomics and Disease Research

In practical applications, the research team used AlphaGenome to study T-cell acute lymphoblastic leukaemia. The AI accurately identified the specific genetic switch manipulations that cause cancer genes to activate. This capability suggests that the root causes of complex diseases can now be identified through computer simulations, potentially reducing the time required for laboratory experimentation. Talc, Asbestos and Cancer: What is the Connection?.

Future of Gene Therapy

While the model is currently limited to research purposes and not yet cleared for clinical use in hospitals, DeepMind has made AlphaGenome open-source. This decision is expected to accelerate global efforts in designing gene therapies and diagnosing rare conditions. By making the code accessible, the company aims to foster a collaborative environment where researchers can enhance the speed of drug discovery and personalised medicine.

(The above story first appeared on LatestLY on Feb 05, 2026 01:18 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).