Researchers at the University of Washington’s Protein Design Laboratory have developed a new variant of diffusion-based novel protein design. produce new antibodies (nature 2025, DOI: 10.1038/s41586-025-09721-5). Researchers are also creating new algorithms. RF antibody, fully usable To researchers.
Researchers in the lab of Nobel Prize winner David Baker have already developed protein minibinders that perform a similar function to antibodies. But Robert Ragotte, one of the authors of the new paper, says that artificial intelligence-designed antibodies have several advantages over minibinders.
Because humans already make their own antibodies, the antibodies are less likely to be immunogenic because the body is more familiar with them. Additionally, while antibodies are already widely used in the pharmaceutical industry, minibinders would be an entirely new approach. However, antibodies have eluded AI protein design programs due to their flexible complementarity-determining region loops. “Flexibility introduces uncertainty into the design process because there is no one true structure,” says Ragotte.
The Baker Lab isn’t the only academic lab to publish new results in recent weeks. In late October, the Boltz team, led by MIT professors Regina Barzilay and Tommi Jaakola, released their latest model.
Boltzendeveloped by PhD student Hannes Stärk, is an all-atom generative model for designing proteins and peptides that bind to a wide range of biomolecular targets.
“The emphasis here is on unresolved problems,” Barzilay said on a call with reporters, who joined right after completing a class on AI and drug discovery. That’s why the Boltz team is focused on finding solutions to goals that are currently unsolvable, she says. The task is Described in preprint It has not yet been peer reviewed.
Chemistry and Engineering News
ISSN 0009-2347
Copyright © 2025 American Chemical Society
