This is the year of artificial intelligence. If on Tuesday Nobel Prize in Physics surprised many by recognizing the artificial neural networks, which have allowed the creation of ChatGPT, this Wednesday the Nobel Prize in Chemistry went to another AI, although this time it was in the pools. The Royal Swedish Academy of Sciences has recognized David Baker of the University of Washington for achieving “the almost impossible feat” of building completely new types of proteins ‘from scratch’, and Demis Hassabis, CEO and co-founder of Google DeepMind , and its director, John Jumper, for the development of Alphafoldan artificial intelligence model capable of predicting the complex structures of proteins at an unprecedented speed.
“We have a great prize today,” said Heiner Linke, president of the Nobel Committee for Chemistry, after announcing the winners. These researchers not only solved a 50-year-old scientific problem but their discoveries “open up enormous possibilities” in many fields of science. Its tools are fundamental to understanding the biology of all living beings on the planet and can promote the development of new medicines or technologies against plastic pollution or resistance to antibiotics. They represent a before and after for science.
Proteins are made up of 20 different amino acids, which can be described as the building blocks of life. In 2003, the American David Baker (Seattle, 1962) managed to use these components to design a new protein that was unlike any other. Since then, his research group has produced one protein after another, including some that can be used as drugs, vaccines, nanomaterials and tiny sensors. «We can create new molecules with surprising functions. “My lab works on proteins that capture and store carbon, assemble into motors, and break down plastic,” told in his day Baker himself told this newspaper.
The second discovery was able to predict the three-dimensional structure of proteins, formed by long chains of amino acids linked together that fold like origami. In 2020, the British Hassabis (London, 1976) and the American John Jumper (Little Rock, Arkansas, 1985) presented an artificial intelligence model called AlphaFold2. With their help, they have been able to predict the structure of virtually all the proteins that researchers have identified, 200 million. Since its breakthrough, AlphaFold2 predictions, freely availablehave been used by more than two million people from 190 countries.
Antibiotics and plastics
Among a host of scientific applications, researchers can now better understand antibiotic resistance and image enzymes that can break down plastic. It has also been applied, among other things, to the understanding of proteins that affect the health of bees and the development of an effective vaccine against malaria. Likewise, it has been used to approach diseases such as leishmaniasis, Chagas disease or leprosy. “We felt like we had changed the world,” Jumper once said in an interview with ABC.
In a telephone connection, Baker acknowledged that, although he had heard his name among possible candidates in recent years, the award for him has been a surprise. «I am very excited. I just want to thank my family and all the people I have worked with all these years and who have made it possible for us to get here.
When asked about the implications of artificial intelligence in the future, Baker stated that “they are tremendous.” «My colleagues and I had been working on protein design for years, but it was not until Hassabis and Jumper’s discovery with artificial intelligence that the power and precision of the tool increased in an incredible way. “I am very excited to think about all the possibilities that this system offers to make a better world, with applications to enhance medicine or technology,” he said.
Accelerate science
«This moment is so incredible, so unreal. “It’s wonderful, such a great honor,” Jumper said excitedly in a press conference after the Nobel announcement. In his opinion, Alphafol allows us to “understand diseases, more people will be healthy with the work we do. We want to make the world a better place and it is wonderful to see every day the work that the scientific community does with this tool.
“Medicine is hard because we don’t understand how the body works, it is a very complicated machine,” Jumper said. “Alphafold can help many scientists around the world understand diseases and fight them faster,” he stressed, while giving as an example the key protein for the transmission of malaria.
Hassabis has expressed himself in a similar way, who stressed that he has dedicated his entire life to artificial intelligence because he is convinced of its “potential to accelerate scientific discoveries.”
AlphaFold’s contributions to science have been widely praised. Among their recognitions are the 2023 Albert Lasker Basic Medical Research Prize or the 2023 Breakthrough Prize in Life Sciences. Baker, Hassabis and Jumper received the Frontiers of Knowledge Award in Biomedicine in 2023. And Hassabis took the Princess of Asturias Award for Scientific and Technical Research in 2022, precisely, among others, together with Geoffrey Hinton, who won the Nobel Prize in Physics on Tuesday.
“I expected it,” Sílvia Osuna, an ICREA researcher at the Institute of Computational Chemistry and Catalysis (IQCC) of the University of Girona, told ABC about the new Nobel Prize. The scientist, who designs enzymes to accelerate chemical reactions and create new drugs to reduce cholesterol or diabetes, considers that the impact of the award-winners’ work “is enormous.” Until 2020, the computational methods that existed to predict the structure of proteins “were not very precise. Alphafold changed everything. It has promoted the development of many tools and what previously took years to achieve is now done in a matter of minutes. This has a great impact on many fields, from medicine to industry,” he points out.
Against malevolent purposes
«Life could not exist without proteins. The fact that we can now predict protein structures and design our own proteins is a great benefit to humanity,” they said from the Swedish academy. However, it’s not all lights. In the same way that computational design can advance new drugs – the first, the SKYCovione vaccine against Covid-19, was approved last year – someone could do the same to implement dangerous biological agents and use them as weapons. Therefore, more than 160 scientists from around the world signed an initiative so that protein design with deep learning tools cannot be used for malicious purposes. Among the promoters of the initiative, Baker, who to avoid it He was committed to “examining and recording all the molecules that are attempted to be manufactured.” Hassabis was betting this Wednesday on “international collaboration” to avoid misuse.
The Nobel Prize in Chemistry is worth 10 million Swedish crowns, that is, 900,000 euros. In 2023, The recognition went to the Frenchman Moungi Bawendi, the American Louis Brus and the Russian Alexei Ekimov for the invention of quantum dots, tiny particles that have multiple applications in the field of nanotechnology, from lighting televisions and LED lamps to tumor tissue surgery.
#Nobel #Prize #Chemistry #Googles #artificial #intelligence #predicts #structure #proteins