Google’s AI research and machine learning company DeepMind announced the release of a new version of its protein prediction tool AlphaFold.
AlphaFold made news several years ago by helping scientists predict the existence of millions of previously undiscovered types of proteins.
This took the combined proteins that scientists had mapped, roughly 170,000, and predicted the existence of over 200 million across 47 organisms.
This protein library was made public to researchers in 2020 in the AlphaFold Protein Structure Database.
Later versions of AlphaFold began to predict and map the structure of the protein molecules, in essence filling out the details of these proteins.
The most recent announcement for the latest version of AlphaFold looks to take these models and begin to predict candidates for molecules that are ripe for binding of small molecule types called ligands.
Ligands are a method for drugs to bind to molecules and provide scientists with a whole new roadmap that shows the way to the creation of drugs that scientists never before dreamed of.
DeepMind and sister Google company Isomorphic Labs are partnering in the creation of the latest round of AlphaFold with the goal of targeting several large global problems.
Early examples they cited include new drugs in the treatment of malaria, and cancer treatment, as well as finding enzymes that can “eat” plastic, effectively disposing of the waste.
DeepMind’s blog post on the subject summed it up saying, “The latest version of AlphaFold’s ability to model such complex systems shows us that AI can help us better understand these types of mechanisms, and accelerate their use for therapeutic applications.”