Researchers using Blue Waters supercomputer make strides in researching treatment for Ebola, other pathogenic diseases

09.26.16 -

by Susan Szuch

Who would have thought that a method that enabled the automatic firing of anti-aircraft guns in World War II would be applicable over 70 years later? This time, though, instead of protecting London's citizens from German warplanes, it’s creating antibodies to protect humans from infectious viruses. Even the method of viral infection is similarly violent to warplanes—viruses like Ebola punch a hole in the surface of a cell to inject genetic material. This method, called smart Monte Carlo or biased random walk, can be explained in terms of evolution: Random mutations occur, but there's a bias toward those mutations that improve survival, since the lethal mutations won't get passed on.

A team of researchers at the University of Illinois at Urbana-Champaign and Stanford University used this method to predict what antibody would most likely pair best with a protein that coats a virus. Their work focuses on two strains of the Ebola virus, and multiple possible mutants of both strains.

"Part of our simulation was to emulate that process in the coat protein. Every time we wanted to try a new coat protein, we threw dice, but we threw loaded dice," team member and senior research scientist at the National Center for Supercomputing Applications (NCSA) Eric Jakobsson, says. "The dice were loaded according to a substitution matrix that comes from aligning zillions (sic) of proteins from corresponding organisms and seeing what happens when one changes."

The use of the substitution matrix to bias the trials of possible antibody sequences reduced the number of possibilities to an extent that testing efficacy with molecular dynamics was manageable, albeit with a substantial allocation on the Blue Waters supercomputer at NCSA. Jakobsson notes that if they continue their research to other viruses with the same method of entry as Ebola, which is most pathogenic viruses, Blue Waters and even bigger machines in the future will be instrumental in assisting the biotechnology industry to produce synthetic antibodies.

"So if we think about gearing this up to provide the computational support for an entire industry, an entire part of the biomedical enterprise that's going to defend us against viral pathogens, yeah, then we'd need all of a Blue Waters machine for that project alone," Jakobsson says.

A key to the work was using published molecular structures from tissue samples from a human survivor and a mouse survivor. The team was able to determine that though the two samples were different strains and had different amino acid building blocks, they had almost exactly the same structures. This structural conservation, along with databases of proteins, allowed them to predict the evolutionary paths the virus would be most likely to take.

In their research they simulated the evolution of the Ebola virus and the most likely mutations of antibody that would combat that evolution. In nature, this is a trial and error process that results in many members of a population dying as it evolves to become resistant. Jakobsson hopes that they may be able to "shortcut" that process and team up with the biotechnology industry to design synthetic antibodies that can be produced on a massive scale.

Ebola isn't the only virus that Jakobsson is concerned about, though. The "big challenge" he hopes the team will tackle: influenza.

Each year, when you receive the flu vaccine, it's based on the strains of the flu that experts anticipate will be circulating during the flu season, according to the Centers for Disease Control. It is educated guesswork but in the last analysis, just that—guesswork. If the best guesses are wrong, the vaccine offers relatively little protection and can be ineffective.

"Everybody who is involved with the flu believes that someday there will be another world-wide pandemic," Jakobsson says. "That someday, we will again be presented with a strain of flu that is highly lethal and highly transmissible, and it's going to kill millions of people—unless we have a better way of dealing with it than we have now."

That better way could be the use of synthetic antibodies, which would have the ability to be quickly redesigned based on the gene sequence of the pandemic strain, thus minimizing guesswork. Jakobsson emphasizes the importance of having a better way to combat influenza, especially in the case of a worldwide outbreak.

"People are moving all around the world, people are carrying viruses with them. It's really in the interests of the developed world to really attack viral infections all over the world, because we are not going to be isolated," he says. "We can't effectively quarantine ourselves. So from self-interest, and also because it's the right thing to do, we should gear up to attack the problem of viral infections around the world."

National Science Foundation

Blue Waters is supported by the National Science Foundation through awards ACI-0725070 and ACI-1238993.