News that Deep Mind’s AlphaFold programme has solved one of the perennial mysteries of biology has filled us with early festive cheer. Even by the standards of biomolecules, proteins are fiendishly complicated. Tiny changes in structure will utterly ruin them, as anyone who has spilled vinegar into milk will tell you. You can’t just bang a load of amino acids together and hope to get a working molecule to cure diabetes or something. The number of permutations and combinations the pesky little molecules can make is just too big.
Or so it was until now. The sheer power of a trained AI system has at last allowed us to predict the outcomes of building these molecules. Here’s Nature to tell you how important this is
An artificial-intelligence (AI) network developed by Google offshoot DeepMind has made enormous progress in solving one of biology’s grandest challenges — determining a protein’s 3D shape from its amino-acid sequence. The breakthrough is likely to transform biology, say scientists, and should aid in drug design. AlphaFold came out on top, by far, in a biennial protein-structure prediction challenge called CASP, short for Critical Assessment of Structure Prediction. AlphaFold’s predictions are comparable in quality to structures determined experimentally by X-ray crystallography or cryo-electron microscopy. “This will change medicine. It will change research. It will change bioengineering. It will change everything,” says evolutionary biologist Andrei Lupas.Nature | 8 min read
Ian Sample of The Guardian tells us a bit about the sorts of applications we might have from this:
DeepMind AI cracks 50-year-old problem of protein folding | DeepMind | The Guardian
Some people will be understandably concerned about systems which are more intelligent than we are running around the planet, and presumably coming into a supermarket near you to buy groceries. To which we say: relax and try to turn down on all this dominance submission stuff! The relationship between humans and their tools has always been symbiotic. The first stone tools let us cut up food better, so soon we didn’t need such big teeth. A short while later the first mobile phones gave you your map, camera, encyclopedia, bank, communications, calculator, games, music and portable drinks mat in your pocket. It’s true we can’t go back to life without them-both our teeth and brains have become too small. But they can’t do without us either – have you ever seen a stone hand axe recharge its own batteries? So it will be with AI, IT, algorithms and all the other potential advances that will change our lives for the better.
#deepmind #proteinfolding #artificialintelligence #medicine