DeepMind’s artificial intelligence, AlphaFold, has made a significant breakthrough in solving the problem of protein folding. This solution could revolutionise the field of biology. Proteins are complex molecules that perform numerous functions in our bodies, and their structures determine their functions. Understanding these structures could lead to advancements in drug discovery and disease understanding.
AlphaFold predicts protein structures with an accuracy comparable to experimental results. It uses two neural networks: one to predict the distance between pairs of amino acids and another to predict the angles between chemical bonds. These predictions are then combined to create a 3D model of the protein.
This technology has already been used to solve the structure of proteins related to COVID-19. It could also be applied in bioengineering, where it could help design proteins with specific functions. Despite its success, AlphaFold is not perfect. It struggles with very large proteins and those that change shape. Nonetheless, this development marks a significant leap forward in computational biology.
Go to source article: https://www.wired.co.uk/article/deepmind-protein-folding