Google DeepMind has launched a new artificial intelligence tool called AlphaGenome, designed to help scientists identify genetic changes that drive disease and support the development of new treatments.
AlphaGenome can analyse up to one million letters of DNA code at a time. It predicts how genetic mutations affect gene regulation, including when genes are switched on, in which cells they act, and how strongly they function.
Many common inherited diseases, such as heart disease, autoimmune disorders, mental health conditions, and several cancers, are linked to changes in gene regulation rather than changes in genes that produce proteins. However, identifying which mutations are responsible has been a major challenge for researchers.
Natasha Latysheva, a researcher at DeepMind, said the tool aims to improve understanding of how the genome works and accelerate research into the biological code that controls life.
The human genome contains around three billion DNA letter pairs. Only about two percent of this code directly instructs cells to make proteins. The remaining portion controls how genes behave by determining where, when, and how much they are activated.
AlphaGenome was trained using publicly available human and mouse genetic data. This allows it to link specific mutations to changes in gene activity across different tissues. The tool can predict how these mutations influence biological processes in organs such as the brain and liver.
Researchers believe AlphaGenome could help identify key genetic regions involved in cancer and other diseases. It may also support the development of gene therapies by enabling scientists to design DNA sequences that activate genes in specific cell types.
Independent experts have described AlphaGenome as a major step forward, while noting that further research is needed before predictions can replace laboratory experiments.
