A new study by an international group of researchers utilizes a novel AI algorithm, PrimateAI-3D, to analyze data from primate DNA and predict potential genetic risks for humans. The scientists sequenced over 800 individual samples from 233 primate species to train the algorithm to recognize benign genetic variants and mutations that could cause diseases. The algorithm achieved impressive results while being able to avoid any genetic bias stemming from white European ancestry. In addition to assisting with health research, the study’s findings can help conserve the populations of other primates. Combining human and non-human primate data can help prioritize the variants that pose the most risk to humans.
The company Illumina collaborated on this research and employed Kyle Farh as co-author. Farh is the Vice President of Artificial Intelligence at Illumina.
The PrimateAI-3D algorithm uses deep-learning language architectures that are similar to those employed with ChatGPT. However, the PrimateAI-3D algorithms model genomic sequences instead of linguistic sequences.
