By james art One of this year’s Nobel Prize winners in physics, Geoffrey Hinton, who pioneered work on the neural networks that underpin artificial intelligence, has warned that machines may one day become smarter than humans. Maybe. But this year’s Nobel Prize in Chemistry honored a real-world example of how AI is helping people today with startling discoveries in protein structure that have far-reaching applications. This is a development worth enjoying.
Proteins are leading actors in biology. As the Nobel Committee pointed out, proteins “control and direct all the chemical reactions that together form the basis of life. Proteins also function as hormones, signaling substances, antibodies, and building blocks of various tissues. In the human body, they are necessary for the structure, function and regulation of tissues and organs. All proteins begin with a chain of up to 20 amino acids strung together in a sequence encoded in DNA. Each chain folds into a unique structure, and these shapes determine how proteins interact with other molecules.
Looking like a tangled ball of string, proteins have a complex and precise design of moving parts that are linked to chemical events and bind to other molecules. Antibodies are proteins produced by the immune system that bind to foreign molecules, including those on the surface of an invading virus, such as the spikes of the coronavirus that causes COVID-19.
In the late 1950s, Cambridge University researchers John Kendrew and Max Perutz successfully used a method called X-ray crystallography to create the first 3D models of proteins. In recognition, they were awarded the Nobel Prize in Chemistry in 1962. Over the ensuing half century, the quest to document protein structures remained laborious and slow. A single protein structure can take a PhD student four or five years to figure out. Before AI, the field’s central repository contained about 185,000 experimentally resolved protein structures.
This year’s Nobel Prize in Chemistry went to three scientists who revolutionized the field. David Baker of the University of Washington has built entirely new types of proteins. Demis Hassabis and John Jumper of DeepMind, a UK-based firm that is part of Alphabet, the parent company of Google, have developed an AI and machine learning model that can predict the structure of proteins by decoding the amino acids that make up each one. The AlphaFold model can do in minutes what once took years.
AlphaFold takes advantage of neural networks that can locate patterns in massive amounts of data. The system was trained on the vast information in databases of all known protein structures and amino acid sequences. AlphaFold has predicted more than 200 million protein structures, or nearly all cataloged proteins known to science, including those in humans, plants, bacteria, animals and other organisms. The AlphaFold Protein Structure Database makes this data freely available.
To design new drugs and vaccines, scientists need to know what a protein looks like or behaves like. The AlphaFold result is a prediction that could accelerate biomedical research. …
The AlphaFold blog tells the story of scientists searching for a better vaccine against malaria, a disease that affects 250 million people a year and causes more than 600,000 deaths. Because malaria is caused by a shape-shifting parasite, vaccine researchers have long struggled to characterize the structure of a single surface protein they must target to interrupt infection. Then AlphaFold’s prediction of the correct structure snapped it into focus. Matthew Higgins of the University of Oxford said the breakthrough helped his team decide which parts of the protein to put into the vaccine, which trains the body’s immune system to detect it and act on it. This helped advance his research from the “basic science stage to the preclinical and clinical development stage.”
Anyone who has used ChatGPT knows that artificial intelligence isn’t always right – and the malaria scientists found that some of the 3D protein visualizations were inaccurate. But AI will only improve over time. Already, AlphaFold’s efforts are expanding to create precise visualizations of how proteins interact with other biomedical structures, such as nucleic acids.
In the coming years, we need to confront the dangers of AI and take precautions. There are undoubtedly risks when powerful technology falls into the hands of malicious players.
But for now, AlphaFold shows that AI can supercharge existing knowledge to benefit humanity. The Nobel committee noted that thanks to this advance, “researchers can now better understand antibiotic resistance and create images of enzymes that can break down plastic.” And there will be more.