Researchers fight Covid with AI


Two papers co-authored by Anima Anandkumar, Bren Professor of Computer Science and Mathematics at Caltech, have been selected as finalists for the 2021 Association for Computing Machinery (ACM) Gordon Bell Special Prize for High Performance Computing-Based COVID-19 Research.

The annual award provides $10,000 from Gordon Bell, a pioneer in high-performance and parallel computing. Ultimately, the 2021 Special Prize went to a team from the Japanese research institute Riken, which simulated how COVID-19 could spread from person to person via aerosolized droplets in a variety of real-world situations.

Both of Anandkumar’s papers studied the coronavirus using artificial intelligence (AI) methods, including those she developed, integrated with large-scale numerical simulations run on supercomputers.

“All six finalists this year had a component in their calculations that used AI,” says Anandkumar. “This has enabled unprecedented understanding of the coronavirus that would not have been possible with traditional tools.”

The research in one of Anandkumar’s finalists, titled “#COVIDisAirborne: AI-Enabled Multiscale Computational Microscopy of Delta SARS-CoV-2 in a Respiratory Aerosol,” used an AI method to model an aerosolized particle of the virus that causes COVID -19 caused, using Oak Ridge National Laboratory’s (ORNL) Summit Supercomputer.

The work, which is being led by a team of UC San Diego researchers including Caltech professor of chemistry Tom Miller; and Zhuoran Qiao, a graduate student, described the interactions of the COVID-19 spike protein and aerosol phase with calcium ions, which are known to play a key role in mucin aggregation in epithelial tissues.

The simulated system, with more than 1 billion atoms, is one of the largest biochemical systems ever modeled at the atomic level. Scientists usually have trouble studying aerosols because of their tiny size.

The Orbnet AI method used in the work was developed by Miller and Anandkumar. It can calculate the quantum mechanical properties of molecules thousands of times faster than conventional methods. Orbnet enabled for the first time the use of quantum mechanical calculations in a large-scale biological system, previously limited to smaller and less complex systems.

The second finalist paper, titled “Intelligent Resolution: Integrating Cryo-EM with AI-driven Multi-resolution Simulations to Observe the SARS-CoV-2 Replication-Transcription Machinery in Action,” was led by a research team from the University of Illinois Urbana-Champaign and contained contributions from Zongyi Li, a graduate student in Anandkumar’s group. The paper used some of the country’s fastest supercomputers to study how the COVID-19 virus reproduces.

The coronavirus uses a tightly coordinated process known as the replication-transcription complex to multiply at high speed when it enters a host’s cells. The team used data from cryo-electron microscopy (cryo-EM), a technique that freezes molecules at lightning speed and uses electrons to create 3D images, to study more closely the molecular machinery involved in this process. Because cryo-EM does not have high enough resolution and time scale to capture important molecular dynamics, the team used a range of AI models to fill the gap with finer-scale simulations. Among these methods was a model called the Graph Neural Operator (GNO) that was used to capture the time-dependent conformational changes in molecular dynamics.

GNO was developed by Anandkumar in collaboration with Caltech’s Andrew Stuart, Bren Professor of Computer Science and Mathematical Sciences; and Kaushik Bhattacharya, Howell N. Tyson, Sr., Professor of Mechanics and Materials Science and Vice Provost. GNO was used as a fast surrogate model to simulate and predict deviations from traditional simulations. In addition, the study of the GNO provided insights into the large conformational changes that occur during the replication process.

“This shows that our GNO method learns useful properties of molecular dynamics during coronavirus replication,” says Anandkumar.

This year, only six finalists were selected for the ACM Gordon Bell Special Prize for High Performance Computing-Based COVID-19 Research.

“It was enormously inspiring to be part of these large interdisciplinary teams that did such a good job of planning and coordinating,” says Anandkumar. “I’ve learned so much about the incredibly complex structure of the coronavirus, the tools needed to take a look at it, and the high-performance computing used to run it at scale. AI has tremendous potential to transform scientific simulations, and these are the first steps. The AI4Science initiative, which I co-founded at Caltech, allows such interdisciplinary collaborations to flourish.”

Anandkumar’s work was funded by the Bren Endowed Chair, and Zongyi Li’s position was funded by the Kortschak Scholars Program and the Caltech DeLogi Fund.


Comments are closed.