Baker and his colleagues created new proteins capable of binding to receptors on cancer cells, trapping metals in solution and binding carbon dioxide for potential use in pulling it out of the atmosphere. In either case, the final predicted proteins can then be produced in the laboratory and tested. The AI keeps bits it deems effective and mutates the rest, steadily progressing towards the goal. The program then generates a virtual protein composed of a random sequence of amino acids and mutates the sequence over and over again, evaluating the effect of each change on the protein’s likely shape and thus function.
SECONDARY STRUCTURE PROTEIN SCAFFOLD SOFTWARE
It gives the software a target for a protein, such as binding to a metal. The second approach, known as limited hallucination, is more open-ended. Just as a word processor’s autocomplete function tries to complete a word after you type a few characters, the AI then draws on its understanding of how proteins fold to fill in additional parts of the protein around the central element. The first, called inpainting, starts like the previous effort, giving the AI a starting point, such as an active site or other key function of a desired protein. Wang, Baker and colleagues have now adapted their AI-powered RoseTTAfold to dream up its own proteins from scratch using two different strategies. “You had to hope there was a good fight,” says Baker. The problem is that the approach only worked when Rosetta identified a suitable scaffold. The software then put the two pieces together and adjusted the combination to make the necessary adjustments. They then had Rosetta scan a database of protein structures they had previously designed and find an existing scaffold that could possibly hold the active site in the correct shape. The team members started by feeding the software an already known piece of what they wanted-a small bit of protein structure, called the binding motif, that is capable of binding to their target. In 2017, for example, researchers led by Wang’s boss, David Baker, a protein designer at the UW, showed that they could use a previously AI-free protein structure prediction software program they had developed, called simply Rosetta, to design potential protein-based drugs which bind to and inactivate molecular targets of the influenza virus and a bacterial toxin. It is one thing to predict how natural proteins might fold it’s another thing to design new ones from scratch. Last year, such companies obtained protein structure prediction software ScienceBreakthrough of the year in 2021. AlphaFold and a similar AI software package called RoseTTAFold also offered thousands of probable structures of different proteins, each bound to a partner with which it pairs inside cells. Last year, an AI program called AlphaFold developed by DeepMind, a sister company of Google, predicted structures for hundreds of thousands of human proteins. The artificial intelligence developed by Wang and his colleagues builds on a number of recent advances in the use of computers to predict the 3D structure of natural proteins from their basic sequence of amino acids. Although researchers have used computers and other means to design new proteins for decades, AI approaches like this are likely to increase successes, Zhang says.
“It’s the perfect use of artificial intelligence,” said Yang Zhang, a protein designer at the University of Michigan, Ann Arbor, who was not involved in the work. The software has already created original compounds for potential use in industrial reactions, cancer treatment and even a vaccine candidate aimed at preventing RSV infections. After watching her son recover quickly from respiratory syncytial virus (RSV), Wang, a postdoctoral fellow at the University of Washington (UW), Seattle, and his colleagues redoubled their efforts, and yesterday revealed in Science a new AI software that can “paint” or “hallucinate” structures for proteins not yet found in nature. Computational biologist Jue Wang was already striving to develop an artificial intelligence (AI) to secrete candidate drugs when he had to rush his 2-year-old son to the hospital with a potentially fatal respiratory infection.
0 kommentar(er)
