DeepMind AI can predict how drugs interact with proteins

DeepMind AI can predict how drugs interact with proteins

Visualisation of a protein binding to a DNA molecule

Science Photo Library/Alamy

An artificial intelligence system can now determine not only how proteins fold but also how they interact with other proteins, drug molecules or DNA. Biochemists and pharmaceutical researchers say the tool has the potential to vastly speed up their work, such as helping to discover new drugs.

Proteins, which play many important roles in living things, are made up of chains of amino acids, but their complex 3D shapes are difficult to predict.

The AI company DeepMind first announced in 2020 that its AlphaFold AI could accurately predict protein structure from amino acid sequences, solving one of the biggest challenges in biology. By the middle of 2021, the company said that it had mapped 98.5 per cent of the proteins in the human body.

Now the latest version, AlphaFold 3, is able to model how proteins, including antibodies, interact with each other, as well as with other biomolecules such as DNA and RNA strands. DeepMind says the accuracy of its predictions is at least 50 per cent higher than existing methods.

Most drug molecules function by binding to specific sites on proteins. AlphaFold 3 could rapidly speed up the development of new drugs by creating a fast way to test how candidate drug molecules interact with proteins in a computer before running lengthy and expensive laboratory tests.

Like earlier versions of AlphaFold, models of proteins or their interactions generated by the latest update aren’t experimentally validated. DeepMind’s chief executive, Demis Hassabis, says AlphaFold 3 only offers predictions, so validation in the lab remains vital – but that research will now be “massively accelerated”.

Julien Bergeron at King’s College London, who wasn’t involved in developing AlphaFold 3 but has been testing it for several months, says it has changed the way his experiments are run. “We can start testing hypotheses before we even go to the lab, and this will really be transformative. I’m pretty much certain that every single structural biology or protein biochemistry research group in the world will immediately adopt this system,” he says.

Keith Willison at Imperial College London says the tool has the potential to streamline large portions of drug discovery and biological research, allowing researchers to focus in on useful molecules that they may never have been able to discover previously.

“Organic chemists used to say the chemical space is larger than the number of atoms in the universe, and we’ll never be able to access even the remotest, tiniest portion of it. But I think these AI techniques are going to be able to access a huge amount of relevant chemical space,” he says.

Matt Higgins at the University of Oxford says the new features in DeepMind’s AI will make a huge difference to biomedical researchers, including in his own work studying host-parasite interactions in malaria.

“While AlphaFold transformed our ability to predict the structures of protein molecules, the protein machines used by our cells rarely work alone,” he says. “AlphaFold 3 brings the new and exciting ability to modify protein molecules with the most common additions or bind them to the most common binding partners found in our bodies and to see what happens.”

Topics:

Read the original article here