Thanks to David Baker, Demis Hassabis, John M. Jumper, and even a little Artificial Intelligence, proteins are less obscure. Yesterday the Nobel Prize for Chemistry, accompanied by a check for 967 thousand euros, went to them. “To understand how life works we need to understand proteins,” explained the Stockholm Committee.
David Baker, a 62-year-old American from Seattle University, was awarded for being the first to understand how to study the structure of proteins and on this basis how to design new ones, useful for obtaining drugs, vaccines, nanomaterials or sensors. Baker has succeeded in a feat that seemed impossible: creating new types of proteins from scratch. In 2003 he presented Top7, a protein that did not exist in nature. It is the year of his consecration.
A Nobel divided and shared with the other two, the American John M. Jumper, only 39 years old, very young among the winners, a physics and mathematics researcher passionate about computer simulations of proteins, and the British Denis Hassabis, 48 years old, inventor of the ‘AlphaFold colleagues at Google Deep Mind in London. And this is where artificial intelligence combined with human intelligence comes into play. Jumper is in the AlphaFold working group and in the past has already developed systems to make calculation programs more efficient. He proposes using something similar to Hassabis. Thus AlphaFold2 was born. It’s the turning point. Their model turns out to be capable of seeing the structure of more than 200 million proteins, virtually all known in the realm of life. Even better: it’s a catalog of sorts. A real revolution, a pioneering and free and accessible system which has allowed over two million researchers spread across 190 countries to study the structure of proteins, thus solving a problem that has lasted for 50 years. A tool capable of shortening research times such as those on antibiotic resistance or the design of enzymes capable of decomposing plastic. «Proteins, the committee explained, form the toolbox of life.
David Baker has succeeded in the almost impossible task of assembling new families of proteins. Hassabis and Jumper developed an artificial intelligence model to solve a half-century-old problem: predicting the complex structures of proteins.”