DeepMind Dismantles AlphaFold Team After Achieving Protein Structure Breakthrough

Google DeepMind's artificial intelligence program, AlphaFold, has made significant strides in predicting protein structure. After eight years and over 200 million predictions, the team behind AlphaFold has been dissolved, with some members leaving the company while others were reassigned to other projects.

The breakup of the AlphaFold team comes as Google DeepMind's CEO, Demis Hassabis, steps down to become chief scientist of Alphabet, Google's parent company. Hassabis had made AlphaFold widely available at no cost, which reportedly caused friction within Google due to its lack of direct revenue.

According to John Moult, a computational biologist at the University of Maryland, AlphaFold 'largely solved' the protein structure-prediction problem, and researchers outside DeepMind have already been carrying forward much of its work. The program's prediction accuracy was an 'amazing achievement in itself,' but it also opened up large areas of science for exploration.

While AlphaFold has made significant progress, many proteins still work as parts of larger molecular machinery, with changing states and binding mechanisms yet to be fully understood. Debora Marks, a computational biologist at Harvard Medical School, notes that 'I'd do exactly the same' as DeepMind, given they were already done.