In a major breakthrough for computational biology, Google DeepMind has released AlphaGenome, an artificial intelligence system designed to decode the mysterious non-coding regions that make up 98% of human DNA.
While only 2% of our genome directly codes for proteins, the remaining 'dark matter' plays a crucial role in regulating gene activity and is often implicated in diseases. AlphaGenome represents the first comprehensive AI model capable of analyzing these complex regulatory regions at unprecedented scale and resolution.
"This is one of the most fundamental problems not just in biology — in all of science," said Pushmeet Kohli, DeepMind's head of AI for science, during a press briefing. The model builds on DeepMind's previous success with AlphaFold, which revolutionized protein structure prediction and shared the Nobel Prize in Chemistry last year.
AlphaGenome's technical capabilities are impressive. It can process DNA sequences up to one million base pairs long while maintaining single-nucleotide resolution, allowing it to predict thousands of molecular properties that characterize gene regulation. In benchmarking tests, it outperformed specialized models in 22 out of 24 sequence prediction tasks and matched or exceeded others in 24 out of 26 variant effect prediction evaluations.
The model has already demonstrated practical applications in disease research. When analyzing mutations found in leukemia patients, AlphaGenome accurately predicted how non-coding variants activated a cancer-driving gene by creating a new binding site for a regulatory protein. "Determining the relevance of different non-coding variants can be extremely challenging, particularly at scale. This tool provides a crucial piece of the puzzle," explained Professor Marc Mansour of University College London.
DeepMind has made AlphaGenome available via API for non-commercial research, with plans for a full release in the future. While the model has limitations—it struggles with very distant DNA interactions and isn't validated for clinical use—it represents a significant step toward understanding how our genome functions and could accelerate discoveries in disease research, synthetic biology, and personalized medicine.