A groundbreaking study from the Australian National University (ANU) suggests artificial intelligence (AI) could soon revolutionise how genetic diseases are diagnosed and treated.
Published in Nature Communications, the research combines AI-driven protein modeling with genome sequencing to better understand how genetic mutations impact human health. The study uses Google DeepMind’s AlphaFold to analyse every possible mutation across the entire human proteome, uncovering why some proteins are more vulnerable to harmful mutations.
Lead researcher Dan Andrews, Associate Professor at ANU, explained that evolution appears to have built resilience into essential proteins, protecting them from destabilising mutations. In contrast, less critical proteins lack this protective trait, making them more susceptible to damage—often playing a larger role in genetic conditions.
“Mutations are constant, like rain on genes,” said Andrews. “While some essential genes remain largely unaffected, others—though less vital—can still cause disease when mutated.”
Researchers from ANU’s John Curtin School of Medical Research and School of Computing say their findings could pave the way for more personalised treatments by identifying which genetic systems are disrupted in individuals.
The study not only aids in understanding monogenic disorders but also holds potential for complex diseases involving multiple mutations. It assesses genetic variations for their functional effects, helping pinpoint which genes are likely malfunctioning.
Looking ahead, the research team plans to develop automated AI tools that can suggest effective treatments based on a patient’s unique genetic and pathology data—bringing personalised medicine closer to reality.