Why won’t AlphaFold have a revolutionary impact on new drug development?

DeepMind announced last month the release of predicted structures for all known proteins, with more than 200 million data in its protein structure database, AlphaFold DB. Will this breakthrough have a revolutionary impact on new drug development? Dr. Derek Lowe, who has worked in several large pharmaceutical companies, gave a negative answer . He points out that the protein structures discovered by AlphaFold DB are known but unknown—the question of why proteins are folded this way will take many years to answer. In addition, the AlphaFold algorithm is based on the ordered known structure, which is not very helpful for the disordered structure. Some disordered proteins suddenly become ordered in the presence of other proteins, and some do not become ordered under any conditions. This feature is critical to its functionality. Parts of protein behavior cannot be explained by structure, clearly beyond the capabilities of AlphaFold.

This article is reproduced from: https://www.solidot.org/story?sid=72396
This site is for inclusion only, and the copyright belongs to the original author.

Leave a Comment