Leading genetic testing startup Ilumina has introduced PrimateAI-3D, an AI algorithm trained on genomic sequencing data from 233 non-human primate species, which predicts disease-causing genetic mutations in patients. The algorithm's performance was demonstrated in clinical cohorts and validated in a population of nearly half a million individuals in the UK Biobank. Illumina plans to make PrimateAI-3D available through Illumina Connected Software.

PrimateAI-3D utilizes deep neural network architectures and is trained on genome sequences to accurately identify disease-causing variants. It leverages natural selection by training on benign genetic variants identified through the sequencing of non-human primate species, allowing improved clinical variant interpretation on a genome-wide scale. The algorithm has shown superior accuracy in identifying disease-causing variants and provides individualized predictions of genetic disease risk.

PrimateAI-3D offers the potential to accelerate personalized genomic medicine by addressing the challenge of identifying disease-causing genetic variants. It enables the identification of rare pathogenic mutations in the general population and improves the accuracy of genetic risk prediction. The algorithm also contributes to the equitable implementation of genetic-based precision medicine for diverse populations.