Google DeepMind has launched AlphaGenome, an AI tool it says can help identify genetic mutations driving disease and point to new treatments.
AlphaGenome predicts how mutations can disrupt gene regulation, affecting when genes switch on, in which cells, and how strongly they are activated.
Many inherited common diseases, including heart disease and autoimmune disorders, as well as mental health problems, have been linked to mutations that affect gene regulation, as have many cancers.
But pinning down which changes matter is often difficult.
The human genome contains about 3bn pairs of letters, the Gs, Ts, Cs and As that make up DNA.
Around 2 per cent carries instructions for making proteins, the body’s building blocks.
The rest helps control gene activity, with signals that influence where, when and how much genes are switched on.
The researchers trained AlphaGenome on public databases of human and mouse genetics, enabling it to learn links between mutations in specific tissues and their impact on gene regulation.
The tool can analyse up to 1m letters of DNA at once and predict how mutations may affect different biological processes.
Natasha Latysheva, a DeepMind researcher, said: “We see AlphaGenome as a tool for understanding what the functional elements in the genome do, which we hope will accelerate our fundamental understanding of the code of life.”
The DeepMind team says the tool could help scientists identify which stretches of genetic code are most important for the development of particular tissues, such as nerve and liver cells, and pinpoint mutations that drive cancer and other diseases.
It could also support gene therapy research by helping researchers design new DNA sequences, for example to switch on a gene in nerve cells but not in muscle cells.
Expert reaction
Carl de Boer, a researcher at the University of British Columbia in Canada, who was not involved in the work, said: “AlphaGenome can identify whether mutations affect genome regulation, which genes are impacted and how, and in what cell types.”
“Ultimately, our goal is to have models that are so good we don’t have to do an experiment to confirm their predictions. While AlphaGenome represents a significant innovation, achieving this goal will require continued work from the scientific community.”
Marc Mansour, a clinical professor of paediatric haemato-oncology at UCL, said the tool marked a “step change” in his work to find genetic drivers for cancer.
Gareth Hawkes, a statistical geneticist at the University of Exeter, added: “The non-coding genome is 98 per cent of our 3bn base pair genome.
“We understand the 2 per cent fairly well, but the fact that we’ve got AlphaGenome that can make predictions of what this other 2.94bn base pair region is doing is a big step forward for us.”
