New Open-Source AI Model Aids Researchers in Interpreting Genetic Variants

The model, developed by an international collaboration, can learn to read and write DNA and understand human language, acting as a translator between these two systems of information.

Carbon-X can understand natural language requests from researchers and perform complex tasks, such as predicting the impact of genetic mutations and identifying regions of the DNA involved in gene regulation.

The system achieves performance close to that of proprietary models, allowing researchers to use it directly on their own infrastructure.