An 8.6-fold speedup in protein folding
Google researchers have developed a method that accelerates protein structure prediction by 8.6 times. Called "surgical epitope excision," the technique focuses structural predictions on the local region where a designed protein is expected to attach, rather than processing the entire target. According to Google, the approach delivered the speedup without sacrificing structural accuracy in its tests.
Molecules that bind to a cancer-linked protein
The team applied the method to FAP (fibroblast activation protein alpha), a molecule associated with solid tumors. Using yeast surface display and fluorescence-activated cell sorting, they screened engineered protein candidates. In laboratory experiments, the strongest designs achieved single-digit nanomolar binding affinity, indicating a tight molecular interaction.
Agents worked with DNA molecules and robots
In a second project, agents improved DNA aptamers, short molecules that fold into shapes capable of attaching to specific proteins. Across three design-and-testing rounds, confirmed on-target hit rates reached 33.5 percent for myoglobin and 42.1 percent for alpha-thrombin. A third project introduced VERID, a digital laboratory model that tracks robotic equipment, liquid volumes and experimental materials so agents can simulate procedures before sending commands.
Still early stage
Google says the three projects remain separate demonstrations and the systems are experimental. The protein binders have not been shown to safely or effectively treat cancer. The company plans to combine its robotic control framework with molecular discovery workflows, keeping human oversight and biosecurity screening throughout the process.
