The most accurate simulation of objects made of tens of millions of atoms has been performed on one of the world’s best supercomputers using artificial intelligence.
Existing simulations that describe in detail how atoms behave, interact and evolve are limited to small molecules due to the computational power required. There are techniques to simulate much larger numbers of atoms over time, but they are based on approximations and are not accurate enough to extract many detailed features of the molecule in question.
Now, Boris Kozinsky at Harvard University and his colleagues have developed a tool called Allegro that can accurately simulate systems with tens of millions of atoms using artificial intelligence.
Kozinsky and his team used the world’s 8th most powerful supercomputer, Perlmutter, to simulate the 44 million atoms involved in HIV’s protein shell. They also simulated other common biological molecules, such as cellulose, a protein lacking in people with haemophilia, and a widespread tobacco plant virus.
“Anything that essentially consists of atoms can be simulated extremely accurately with these methods, and now also on a large scale,” says Kozinsky. “This is one demonstration, but by no means limited to this domain.” The system could also be used for many problems in materials science, such as investigating batteries, catalysis and semiconductors, he says.
To simulate such large numbers of particles, the researchers used a kind of AI called a neural network to calculate interactions between atoms that were symmetrical from every angle, a principle called equivariance.
“If you develop networks that encompass these symmetries very fundamentally, you get these big improvements in accuracy and other properties that we care about, like the stability of simulations, or how fast the machine learning model learns when you learn it with more data, says team member Albert Musaelianalso at Harvard.
“This is a tour de force in programming and demonstrating that these machine-learned capabilities are now scalable,” says Gabor Csanyi at the University of Cambridge.
Simulating such biological molecules, however, is more of a demonstration that the tool works for large systems than a practical boost for researchers, as biochemists already have sufficiently accurate tools that can be performed much more quickly, he says. Where it could be useful is for materials with many atoms that experience shocks and extreme forces on very short time scales, such as in planetary nuclei, says Csányi.
Subjects:
