San Diego Lab Takes ‘Open Science’ Approach To Tracking Zika’s Arrival In US
Wednesday, May 24, 2017
Photo by Nicholas McVicker
San Diego scientists are out with a new study that traces the spread of Zika into the United States. Their research, published Wednesday in the journal Nature, shows that the virus arrived in Florida multiple times, likely through busy travel routes connecting Miami with popular destinations in the Caribbean.
But the thing is, this new study is not actually all that new. The researchers have known pretty much everything in the paper for months. And they made their findings public long before this week's publication.
They say their approach is part of a larger trend toward "open science," which aims to make results available quickly — before they are formally peer-reviewed and published in a scientific journal.
The research relied on gene sequencing machines like the ones humming away in a lab at the Scripps Research Institute in La Jolla. "This is where all the magic happens," said Scripps postdoctoral fellow Nathan Grubaugh, the first author on the Nature study.
He and his colleagues used machines like this to decode dozens of Zika virus genomes. Their samples came from Florida, where Zika — the virus linked with severe birth defects in countries like Brazil — began showing up in mosquitoes and humans last year.
A new Zika study led by San Diego scientists is not actually all that new. The researchers have known pretty much everything in the paper for months, and they made their findings public long before the study was published in a journal.
The research team's goal was to find out when, how and why this virus first came to the U.S. And they knew these genomes could reveal a lot.
Grubaugh opened his laptop and pulled up a file full of sequencing data. It looked like a big Excel spreadsheet, overflowing with A's, C's, G's and T's. He pointed to one highlighted column.
"Some of the viruses have a G there and some of them have an A," he said.
Those small genetic differences tell an important story. They are mutations that actually show how the Zika virus changed over time as it spread to new locations.
Grubaugh explained, "If you just give me an unknown Zika virus and ask me to sequence it, I can quickly tell if it came from Florida or some other place. And likely with pretty decent resolution of the country it likely came from, because of this relatedness."
Using these genomes, the researchers could recreate a kind of family tree for each sample. That information revealed that Zika mostly came to Florida not from South America, but from the Caribbean.
Grubaugh and his co-authors conclude that Zika came to Florida at least four separate times — and perhaps as many as 40 times — possibly through cruise lines and air travel routes linking Miami and islands throughout the Caribbean.
Grubaugh said more research is needed to pinpoint which islands have been the most significant exporters of Zika to Florida. But he said, "It could be the Bahamas or the U.S. Virgin Islands or such. It makes sense. There's a lot of people who go there to vacation and come back. Or Jamaica or Puerto Rico, where you have a lot of family members."
These findings are just now coming out in a scientific journal. But a very similar version of this study has actually been publicly available for months. The day the researchers submitted their study to journals, they also posted it to a website called BioRxiv, which hosts "preprints," or manuscripts that scientists upload to the Internet before they're accepted for publication in a traditional journal.
"This was posted on February 3rd, months before it will actually hit mainstream press," Grubaugh said. Plus, long before that, the researchers were posting their raw data to other websites in close to real time.
"We can go from sample to having data online in five to seven days," said Grubaugh. "And that's, a lot of times, what we shoot for."
Grubaugh and his colleagues are among the many scientists now choosing to get their research out in the open as quickly as possible.
They say this approach can make science more transparent and nimble, given that waiting to publish in a journal can keep findings hidden behind closed lab doors for many months. And because other researchers can scrutinize and comment on their early findings, they say this approach can actually enhance peer review.
Grubaugh said on a research effort like this, it would not have been right to stick with the more traditional approach of closely guarding data until publication.
"There's an epidemic happening," he said. "And it's wrong in a way to withhold information from the rest of the world that could be beneficial to helping to slow down or to prevent Zika virus infections. Or for other researchers to use what you find to apply to their own work."
Grubaugh is a postdoc in Kristian Andersen's lab at Scripps. Andersen, the lead author on the Nature paper, also feels it is wrong to withhold access to ongoing research about a significant public health threat.
Last week, he tweeted about another recently published Zika study, saying, "This paper is everything that's wrong with publishing. Paywalled, no data available, no preprint. Sad!"
The Andersen lab's website includes a section where lab members post their data. It is labeled "secrets."
"It's just a joke that we make all our secrets publicly available," said Grubaugh.
But the joke gets at a serious point. In some ways, scientists have an incentive to keep secrets. They can advance their careers by being the first to publish big new findings in a top journal. Some might worry that if they share their data too widely, competing scientists might seize on that data to scoop them.
Grubaugh said that is an old-school mindset, and his team's experience shows that just the opposite can be true.
By posting their early results to sites like Twitter, GitHub and nextstrain.org, they were able to form helpful collaborations with researchers at other institutions like the U.S. Army Medical Research Institute of Infectious Diseases and the Florida Department of Health, who ended up being co-authors on the paper.
"Instead of competing, we decided, let's combine our datasets," Grubaugh said. "A lot of it wasn't us seeking help. It was help coming to us because we were open about what we were doing."
One thing most scientists still avoid before publishing in a journal is talking with the media about their early results. Which is why you end up with stories like this one, talking about new studies that are technically not that new. In fact, the next big "new study" you read about could already be online now — if you know where to look.
To view PDF documents, Download Acrobat Reader.