Artificial Intelligence (AI) is a tool, which enables people to analyse huge pools of data very quickly and with high degree of accuracy. AI could speed up a lot of our scientific work. For example, researchers at Chalmers University of Technology, Sweden, have found that AI is able to generate new proteins that could speed up drug development.
Proteins are biologically important for all living cells. They can build, modify or remove other molecules. This is very important and also the reason why proteins are so important for drug development. In fact, pretty much all the most effective cancer, diabetes and other medicine are protein-based. We could use proteins in our medicine development even more extensively if we could develop them faster.
Proteins are not engineered through introducing random mutations to protein sequences. Protein activity declines with each of those mutations, which makes protein development cost a lot of time and money. And that’s where the new AI-based protein engineering method comes in.
ProteinGAN, as this AI-based method is called, allows scientists to go from computer design to working protein in just a few weeks and that is why it is called one of the biggest pharmaceutical breakthroughs in recent times. Not only does it speed up the process, but it also helps save money, which is very important in scientific research, because funding is always limited. But how does it work?
Well, essentially ProteinGAN is fed with a large amount of data from well-studied proteins. AI system analyses that data and models new proteins based on it. Another part of the system recognizes fake proteins. These parts of the AI talk between themselves until something good is created. Of course, at the time it only exists as a computer model, but scientists can then bring it to life in a laboratory setting.
Interestingly, the same kind of thinking could be applied to industrial processes. Martin Engqvist, one of the scientists involved in designing the experiments to test the AI synthesised proteins, said: “Accelerating the rate at which we engineer proteins is very important for driving down development costs for enzyme catalysts. This is the key for realising environmentally sustainable industrial processes and consumer products, and our AI model, as well as future models, will enable that. Our work is a vital contribution in that context”.
Now scientists will try to explore how this technology could be applied to tune specific properties of those proteins. At the end there should be a powerful tool, which could speed up drug development and testing. Maybe someday scientists will be able to simply request a protein of specific properties and the AI will do the rest.
Source: Chalmers University of Technology