Google DeepMind demonstrated on October 15, 2026, that its Antigravity platform can link more than 30 scientific databases to accelerate life-sciences research. In one demonstration, the ProtNLM model analyzed millions of unannotated UniProt entries to identify a new antimicrobial peptide candidate. The peptide showed promising activity against Streptococcus pseudopneumoniae, was evaluated in silico with AlphaFold, and was used to produce a wet-lab experimental plan. That plan remains in the validation phase, while Antigravity can complete complex workflows in minutes instead of days and generate publication-ready outputs without requiring scientists to write code. The platform builds on Science Skills, a suite of more than 11 proteomics and cheminformatics tools introduced by Google in May 2026.
Source: Read the original article
