Whether it’s monitoring vehicles for lane drift, recognizing human faces or voices to unlock a phone, or recommending e-commerce products or streaming movies — most people know that AI is watching and interacting with us more than ever before. What is not so well-known is how much AI has also established itself as a crucial tool in academics, particularly for scientific discovery.
An interdisciplinary team of chemists, physicists, and computer scientists from the University of Warwick, the Technical University of Berlin, and the University of Luxembourg have developed a deep learning algorithm that leverages fundamental quantum mechanics equations to accurately predict the quantum mechanical wave functions of molecules. This provides a probability distribution for the value of each observable of the quantum system, from which additional ground-truth properties of the molecules can be derived.
The research team say the study “opens promising avenues to perform inverse design of molecular structures for targeting electronic property optimization and a clear path towards increased synergy of machine learning and quantum chemistry,”
The team was assembled through a UCLA fellowship program on machine learning in quantum physics and has been working together for three years.
Paper co-author Reinhard Maurer from the Department of Chemistry at the University of Warwick says the novel research undertaking “required computer science know-how to develop an artificial intelligence algorithm flexible enough to capture the shape and behaviour of wave functions, but also chemistry and physics know-how to process and represent quantum chemical data in a form that is manageable for the algorithm.”
To predict the quantum mechanical wave functions, the algorithm first learnt to solve fundamental equations of quantum mechanics.
Solving these equations in the conventional way requires massive high-performance computing resources, and can often take months of very pricey computing time. For this reason, compute is typically the bottleneck in the design of new purpose-built molecules for medical and industrial applications. The newly developed algorithm however is able to supply accurate predictions within seconds on just a laptop or smartphone. This output can be applied to drastically speed up future simulation efforts for the design of drug molecules or new materials.
The study is “an important progress as it shows that AI methods can efficiently perform the most difficult aspects of quantum molecular simulations,” says paper co-author Klaus Robert-Muller from the Institute of Software Engineering and Theoretical Computer Science at the Technical University of Berlin. “Within the next few years, AI methods will establish themselves as essential part of the discovery process in computational chemistry and molecular physics.”
The paper Unifying machine learning and quantum chemistry with a deep neural network for molecular wavefunctions was released in June on arXiv and published in the November Nature Communications.
Journalist: Yuan Yuan | Editor: Michael Sarazen