A multi-institution research collaboration in China has built a system that generates functional proteins directly from natural-language prompts describing the desired function, rather than requiring detailed structural constraints like amino acid positions. Separately, Hugging Face announced it is hosting a similar tool for designing DNA sequences from text prompts. The protein tool's success rate is limited, producing many non-functional designs alongside working ones.
arstechnica.com
· 2026-10-27
Researchers released Pinal, a system that designs functional proteins from natural-language descriptions, under an Apache 2.0 license, with code on GitHub and a public web server. The team withheld full release of its large-scale training corpus and pipeline, citing biosecurity concerns, while publishing inference code, evaluation code, model weights and usage instructions instead. The training data came from AFDB, which excludes viral proteomes, so the developers say Pinal was not built for viral protein design.
nature.com
· 2026-10-21
Scientists report using generative protein design to build entirely new protein scaffolds that bind metal polypyridine photocatalysts, which previously lacked natural protein partners capable of enantioselective control. Directed evolution refined these designed proteins into artificial metalloenzymes that perform efficient, highly enantioselective light-driven reactions, including with cheap racemic metal complexes, and support high catalyst turnover and recycling. Spectroscopy showed the protein environment extends the photocatalyst's excited-state lifetime and positions substrates to control stereochemistry.
nature.com
· 2026-10-21
Researchers used AI protein-design tools to build new enzymes by keeping the catalytic core from two existing protease types and replacing the surrounding protein structure. The engineered enzymes can be tuned to cleave particular target proteins, and the team tested this by using them to cut proteins linked to neurological diseases.
arstechnica.com
· 2026-10-14
The acknowledgments section of a paper on computational design of metalloproteases thanks staff and collaborators at the Institute for Protein Design for computational and laboratory support, along with researchers who provided reagents and feedback. It also credits beamline facilities at the Advanced Photon Source, National Synchrotron Light Source II and the Center for BioMolecular Structure, funded by the US Department of Energy and National Institutes of Health, for structural data collection.
nature.com
· 2026-10-14
Google DeepMind researchers in London have built a watermarking system, described in Nature, that embeds a hidden statistical signal into both the amino-acid sequence and 3D structure of AI-designed proteins. The tag, called SynthIDBio, is meant to mark proteins as machine-generated without impairing their biological function, allowing them to be identified later in databases.
nature.com
· 2026-09-30
Researchers have extended Google's SynthID-text watermarking system to ProteinMPNN, a widely used protein sequence design model, creating a method called SynthIDBio-sequence. The approach embeds a detectable signal during the autoregressive sampling of amino acids, using a random seed, scoring function and tournament sampling algorithm similar to the text version. Because protein design typically uses low-temperature sampling that limits randomness, the team also developed a distortionary variant with an added filter to preserve detectability.
nature.com
· 2026-09-30
David Baker, who shared the 2024 Nobel Prize in Chemistry for computational protein design, is leading a new initiative called AI BioDesign, backed by the Allen Institute, University of Washington and Fred Hutchinson Cancer Center. The project pairs AI models with large-scale lab experiments to design molecules and biological functions that don't exist in nature but are chemically feasible, aiming to build public databases and tools for future medicines and technologies.
wired.com
· 2026-09-18