Scientists Hope To Use Primate DNA To Train AI Models For Analysis Of Human Genetic Diseases
Genetic diseases are still quite difficult medical problems to solve, but the advancement of AI technology may bring the answer.
Recent studies have tried to use primate DNA to train AI models, and future attempts will be used to analyze human genetic diseases.
Scientists first used 233 non-human primate species to sequence the genes of more than 800 individuals, ranging from lemurs to gorillas, and then used an algorithm called PrimateAI-3D to analyze and make models.
Its architecture is similar to ChatGPT, which is a deep learning system that can eventually be used to identify benign genetic mutations and find mutations that may cause diseases through the method of elimination.
They tried to use this model to test human data, using more than 400 voluntary DNA data from UK Biobank, showing that it has a very good performance in predicting the risk of human genetic diseases, which is more accurate than existing technology, and is not limited to white European ancestry.
One of the participating scientists, Kyle Farh, said that the data from humans is not enough to make effective models, so to truly understand human genes, the genes of more species need to be analyzed.
If the genetic variation exists in other primate genes, the chance of causing the disease is only 1{9fa2ae831048424756faedea87f44870cd2ea6d09bcfab62eb5dabc6b233e172}. In addition to helping to predict human genetic diseases, it can also help protect the ecology of other primates.
