The artificial intelligence model known as “AlphaFold 3” has been developed by DeepMind, a subsidiary of Google’s Alphabet company. This model goes beyond simply calculating proteins and is able to make highly accurate predictions for how proteins interact with other biomolecules within human cells. Isomorphic Labs, a London-based subsidiary of DeepMind specializing in artificial intelligence in drug research, played a crucial role in the development of this model.
Previously, DeepMind created AlphaFold 2, a program capable of calculating the three-dimensional shapes of proteins. Understanding the shape and function of proteins is crucial for comprehending biological processes in the body and diseases. In 2020, DeepMind made a groundbreaking advancement in predicting protein structures, leading to significant discoveries in areas like malaria vaccines, cancer treatment, and enzyme design using AlphaFold 2.
In an article published in the journal Nature, DeepMind introduced AlphaFold 3 as an artificial intelligence system that can predict the structure and interactions of all life molecules with unparalleled accuracy. The system has shown significant improvements in predicting protein interactions with other molecules compared to existing methods, doubling the accuracy for some essential interaction categories.
DeepMind also unveiled the “AlphaFold Server,” a free tool that offers non-commercial researchers access to the capabilities of AlphaFold 3 for generating models of biological structures quickly and easily. Isomorphic Labs is collaborating with pharmaceutical companies to leverage the potential of AlphaFold 3 for drug development.
Overall, AlphaFold 3 represents a significant advancement in the field of protein structure prediction, opening up new possibilities for understanding biological processes and developing novel treatments.