Stanford Lab Releases Proto for Combining AI Biology Models

A research laboratory at Stanford, in collaboration with the Arc Institute, has introduced Proto, an innovative open-source framework. This new tool addresses a long-standing challenge in AI biology by allowing researchers to integrate and run multiple AI models together, rather than individually. Previously, incompatible software and data formats made combining these models, which are used for designing proteins, DNA, and RNA, a complex and time-consuming process. Proto streamlines this by providing a unified language, enabling users to specify a desired molecular function, and the framework then orchestrates the appropriate models to achieve the design. This capability significantly reduces the experimental burden; for instance, engineering cell-specific genetic switches now requires testing dozens of candidates instead of thousands. The framework is also designed to be programmable by AI agents, potentially becoming a standard interface for future biology models.