Exeter researcher pioneers AlphaGenome Atlas, identifying genetic regions that could link to disease

The University of Exeter has collaborated with Google DeepMind, becoming among the first researchers in the world to test a new platform which contain predictions for the effects of 9 billion genetic variants – every single letter change in the genome.

Exeter academics tested AlphaGenome Atlas, the most comprehensive catalogue to date to predict how genetic changes impact molecular biology. Fundamentally, AlphaGenome Atlas assigns each genetic variant a number, and the higher this number, the more likely it is that the variant causes disease, allowing researchers to prioritise where to focus their attention within 9 billion variants.

Dr Gareth Hawkes, a Medical Research Council fellow at the University of Exeter Medical School, applied AlphaGenome Atlas to whole-genome data from over 54,000 UK Biobank participants, which made elusive genetic signals more identifiable. He said: “The smaller we can make the haystack, the more likely we are to find the needle.” By grouping rare variants based on their predicted molecular effects, Dr Hawkes uncovered 22 per cent more non-coding genetic associations, which would otherwise not have been detectable in the statistical noise. This enabled Dr Hawkes to pinpoint specific regulatory variants driving the abundance of critical proteins circulating in the human body, including PLA2G7 (linked to aging) and EGLN1 (a vital cellular oxygen sensor). Taking this approach even further, Dr Hawkes used AlphaGenome Atlas to look at how hundreds of millions of non-coding variants in the UK Biobank might be linked to body mass index. By focusing on the 1 per cent of non-coding variants which Atlas predicts to be most impactful, he identified genetic regions which could help direct the next stage of targeted research into this trait.

Dr Hawkes also led the first “Unmasking the Genome” workshop, hosted last week at the University of Exeter in partnership with the UK Functional Genomics Initiative and Google DeepMind. The workshop was a highly immersive integration of cutting-edge whole-genome sequencing (WGS), AI-driven variant interpretation, including hands-on training with AlphaGenome and practical functional validation. As part of the workshop, Exeter’s researchers gave workshops in rare disease genetics, population-scale rare variant association testing, and genetic variant interpretation.