Rutgers researchers using power of Web to help stop Zika

A global project led by researchers at Rutgers New Jersey Medical School in Newark is using a crowd-sourced supercomputer network to test potential cures for the Zika virus, the infection expected to spread into the southern United States this summer.

You can help: Just download an app to allow the IBM-sponsored “world community grid” to use your computer when it’s idle.

The massive supercomputer, created by donations of such excess capacity by hundreds of thousands of computer users, will enable speedy screening of millions of potential compounds to identify promising drug candidates for testing in the laboratory. It’s already been harnessed to winnow potential new treatments for malaria, HIV/AIDS and Ebola.

“We’re trying to make finding drugs more efficient, in terms of time and money,” said Joel Freundlich, who heads a chemical biology laboratory at the medical school and is a key consultant on the project.

“Instead of having to wait a number of years, even decades potentially, to test all these compounds in order to find a few that could form the basis of antiviral drugs to treat Zika, we will perform these initial tests in a matter of months,” said Alex Perryman, a member of Freundlich’s lab group and a co-leader of the research. Perryman pioneered the use of supercomputers in the quest for drugs against AIDS and malaria.

Perryman’s method uses a computer model of the virus, detailed to the level of individual atoms, to test how various compounds fit in or interact with it. In contrast, a pharmaceutical lab typically tests thousands of compounds against a pathogen by actually applying the test material to a tiny amount of the virus or bacteria in a culture.

Picture the Zika virus as covered by a number of different locks. The compounds to be tested are like keys. A key that fits in a lock can jam the virus’ ability to reproduce or to sicken its host, and thus be a potential candidate for further study. The project also will use atomic models of related viruses, such as dengue fever and West Nile, in addition to Zika, to test for possible broad-spectrum treatments for the family of related diseases.

“I’ve created 200 million initial docking jobs to evaluate,” Perryman said this week, a day before the OpenZikaproject was to begin.

The “keys” come from libraries of existing drugs as well as simpler compounds. All are in the public domain and made available through the National Institutes of Health and other sources. The results generated from the project also will be available to scientists who ask for them.

 

Zika’s damage

Zika, transmitted by the bite of an infected mosquito, has spread through Latin America and Puerto Rico and affected hundreds of travelers coming to the United States. The virus destroys brain tissue in some babies born to infected mothers, causing small brains and heads. While most people infected with the virus show no symptoms, in rare cases it can cause neurological problems, such as Guillain-Barré syndrome. President Obama asked Congress for $1.9 billion to fight the virus, an amount that has been reduced by the House of Representatives and is being debated.

The crowd-sourced computer program uses idle computing power that would otherwise go to waste. While it is running, a simplified graphic showing some part of the calculation under way appears on the screen. If the computer needs to be used for other purposes, the program disengages.

The security of volunteers’ computers has never been compromised, IBM spokesman Ari Fishkind said.

Security “has been our No. 1 concern since the very beginning, 12 years ago,” Viktors Berstis, IBM’s lead scientist for the grid, said by telephone Wednesday. “We do extensive software audits of all the software involved, so there are no vulnerabilities or Trojan horses or viruses that could cause a problem.”

The scope of the grid’s contribution is huge: The work processed in the past 12 years is the equivalent of a single computer running for 1.2 million years. More than 3 million computers and mobile devices used by nearly 750,000 people and 470 institutions across 80 countries have been used for projects valued at more than $500 million.

Some 2 million scientific calculations are performed per day, Fishkind said. Among the breakthroughs has been the identification of seven drug candidates for treatment of neuroblastoma, a type of cancer affecting infants and young children.

The grid is both larger than government or academic supercomputers, and more environmentally friendly. If all of the computer power being used were assembled in one place, a lot of electricity would be required to cool the machines. Instead, the dispersed machines generate less heat.

Those who join the network may participate in a project forum and ask questions of the scientists; Perryman said he relished responding to such inquiries in previous projects.

“We think this is a great way to raise awareness of public health,” said Freundlich, who has visited his children’s schools to talk about the research.

 

Faster testing

The Go Fight Against Malaria project, which Perryman designed and developed in his previous post at the Scripps Research Institute, tested more than 1 billion combinations of “locks and keys” in less than two years using the IBM community grid, which included 2 million computers at that time. Now, it has increased to 3 million computers, so testing should proceed more quickly, Perryman said.

“I think we’ll easily do over 1 billion [lock-and-key combinations] in less than two years,” he said. As the process narrows the field of potentially useful substances, they will be tested in a laboratory. Years of additional work and the contributions of other scientists will be needed to enhance the potency and specificity of the different agents and reduce their toxicity before they can be used to treat people.

For now, scientists are donating their time, and IBM’s foundation is supporting the use of the supercomputer. OpenZika is led by Carolina Horta Andrade, the principal investigator and a professor of pharmacy at the Federal University of Goias in Brazil. Perryman is co-principal investigator along with Sean Ekins, the CEO of Collaborations Pharmaceuticals.

“It’s like trying to climb a mountain” with multiple approach routes, Freundlich said. “We’re looking for a foothold.”