The physics of a deadly virus

Immune response research hints at answers to fighting deadly coronavirus

By Ignacio Lobos

When a new and deadly coronavirus began to sweep across the world earlier this year, researchers from the UW’s Department of Physics, the University of Hong Kong and other institutions quickly assembled a team to learn how B cells — a central player in adaptive immunity — were engaging this enemy.

With about 10 billion B cells circulating in the human body at any given time constantly searching for invaders, that’s no easy task, said the UW’s Zach Montague, one of the team’s principal collaborators. Each one of these B cells is unique, with a Y-shaped receptor capable of coupling to a specific antigen from an invading virus or bacteria.

But that’s just in one person, Montague said. Each person makes very different and diverse antibodies, even against the same virus. Finding a B cell that can be enlisted to fight COVID-19 across an entire population is literally like finding a single unique grain of sand among the grains that cover all the beaches in the world.

Zachary Montague PhD Student, Department of Physics, University of Washington

Zachary Montague PhD Student, Department of Physics, University of Washington
(Photo credit: Gregory Evans)

Our results provided important insights into COVID-19, and our findings could enhance research and the development of effective therapies and vaccines against this disease.
Zachary MontaguePh.D. Student, Department of Physics, University of Washington

Yet that’s just what Montague and his colleagues set out to do — using an interdisciplinary approach that blended biology and physics, and boosted by the computing power of Hyak, the UW’s supercomputer.

In July, Montague’s group of experimentalists and theorists found 34 potential “grains of sand” in the B-cell repertoire of 19 COVID-19 patients from Hong Kong — candidate reactive antibodies that appeared to have expanded in response to the infection and were present in more than one person.

By dramatically narrowing the field to a few dozen antibodies, the team opened the door to more targeted studies by medical researchers. And that knowledge could ultimately lead to powerful new ways to enlist the body’s immune system in fighting the disease, including the production of monoclonal antibodies, drugs based on B-cell antibodies, to fight COVID-19.

“This is a terrible disease and we have to do everything we can to find every possible way to beat it,” Montague said. “Our results provided important insights into COVID-19, and our findings could enhance research and the development of effective therapies and vaccines against this disease.”

When speed matters

The findings were made in an incredibly short amount of time, thanks to the intertwining of modern biology and physics, collaboration across borders, and sophisticated tools such as Hyak, the UW’s supercomputer, which can handle vast amounts of calculations quickly and accurately.

“When so many lives depend on the outcome of research, speed becomes an essential factor of our work,” Montague said.

Hyak, which is operated by UW Information Technology (UW-IT), was declared an essential research facility early during the pandemic to ensure critical research like this could continue without interruptions. Currently, other pandemic-related research on Hyak is under way, including a collaboration between the Department of Bioengineering and UW Medicine. That study is looking at using artificial intelligence techniques to help predict which COVID-19 patients are at highest risk for heart complications from the illness.

When so many lives depend on the outcome of research, speed becomes an essential factor of our work.
Zachary MontagueUW Department of Physics

“Hyak remains an impressive resource for research at the UW,” said Nam Pho, UW-IT’s Director of Research Computing. Pho manages Hyak and provides training and education to make high-performance computing more accessible to the broader University community.

“We were thrilled that Montague and his colleagues mobilized so quickly, got their hands on the data and conducted valuable analysis using Hyak,” Pho said.

At the interface of statistical physics and biology

Montague is a Ph.D. candidate who specializes in the study of B cells and their interactions with HIV and now COVID-19, and conducts his research as a member of Assistant Professor Armita Nourmohammad’s statistical physics of evolving systems group. He relies on Hyak to construct and infer predictive models for the evolution of B-cell receptors in response to evolving pathogens.

By using mathematical techniques and building novel algorithms, statistical physicists such as Montague seek to understand how the various components in a person’s immune system work together or independently to mount a defense against an invader.

“The problem-solving mindset of a physicist is very useful in many fields,” said Montague, who studied particle physics before coming to the UW.

“When I started looking at other fields for my Ph.D., I found out about Armita and her work with immunology,” he said. “Using physics to study the immune system makes for a very versatile and exciting field. I fell in love with the field of immunology, and I was completely enthralled by B cells and their role in an extremely complex immune system.”

The interactions between B cells and COVID-19 offer a perfect working model for statistical physics and modern biology, said Nourmohammad, a corresponding author of the B-cell/COVID-19 study.

“We were literally looking for the needle in the haystack — but even more tricky because we are dealing with complex and dynamic systems,” Nourmohammad said.

“Things are happening very fast with COVID-19, and it’s quite amazing to have resources such as Hyak to make this research happen. The work we accomplished so quickly was a huge team effort, and we couldn’t have done it without Hyak,” she said.

Statistical physics studies the behavior of large collections of interacting objects, such as atoms and molecules. Its reach has expanded into other areas of study, including biology, which offers a nearly limitless world of dynamic systems.

“I’m a physicist but I also rely on the methods and software tools that you find in bioinformatics to understand biological data,” Montague said.

“I work with nucleotide data (genome, gene and transcript sequence data), which means I have to do all the work that biology people do, such as align sequences, match genes, and perform many other processes that take a lot of time and computing resources.

“Hyak allows me to do all this heavy work in a pragmatic amount of time. What used to be weeks and weeks of computation work, I can do in hours or minutes in Hyak,” he said.

Finding a needle in a haystack

To shape their study, the team started with some basic questions: Which B cells recognize SARS-CoV-2, the novel coronavirus more commonly known as COVID-19, and react to it among the many people infected by the disease? And how many of these B cells capable of mounting an immune response specifically against COVID-19 do we have in common?

To find the answers, researchers in Hong Kong earlier this year collected biological data from 19 patients infected by COVID-19 — at different stages and severity of the disease, from mild to severe —and three healthy individuals.

Using the latest gene sequencing tools, they turned plasma samples into nucleotide data and sent it to Montague, who used Hyak to analyze it.

Things are happening very fast with COVID-19, and it’s quite amazing to have resources such as Hyak to make this research happen.
Armita NourmohammadAssistant Professor, UW Department of Physics

“Hyak has been super useful for our work, speeding up the time to get results. I can ask it to run an algorithm at 3 a.m. and wake up and have the results by 9 a.m.,” Montague said.

In all, the team found 34 unique clonal lineages — B cells with a specific receptor that appeared to have expanded throughout the COVID-19 infection within the volunteer pool. Theoretically, if B cells are cloning themselves, it could mean they’re actively engaging the virus. However, the data needs to be tested and confirmed in a lab to determine if they’re reactive to the disease, Nourmohammad said.

Montague said the team also saw activation of antibodies with cross-reactivity to SARS-CoV, which caused severe acute respiratory syndrome (SARS) in several countries in 2003, and SARS-Cov-2, a significant observation that needs further study.

“There’s a tremendous amount of variability in how we each independently react to disease,” Montague said. “Understanding how our immune system adapts to fight a virus like COVID-19 may help us with the appearance of other novel viruses in the future.”