Researchers at Constructor University and Constructor Labs have developed BiteNetI, a deep-learning model that locates the binding sites of 14 biologically important ion types directly in three-dimensional protein structures. The model needs only several seconds per structure and reaches two- to threefold higher accuracy than most existing predictors, including Google DeepMind's AlphaFold 3. The study, authored by Igor Kozlovskii and Petr Popov, has been published in the journal Communications Biology. The tool provides an open-access platform that could help accelerate drug discovery and the understanding of protein function.

Researchers at Constructor University and Constructor Labs have developed BiteNetI, a deep-learning model that locates the binding sites of 14 biologically important ion types directly in three-dimensional protein structures. The model needs only several seconds per structure and reaches two- to threefold higher accuracy than most existing predictors, including Google DeepMind’s AlphaFold 3. The study, authored by Igor Kozlovskii and Petr Popov, has been published in the journal Communications Biology. The tool provides an open-access platform that could help accelerate drug discovery and the understanding of protein function.

From Biology News - Evolution, Cell theory, Gene theory, Microbiology, Biotechnology via This RSS Feed.