A bold experiment in open science

by Daniel McCabe, BA'89

The Montreal Neurological Institute will soon make all of its research data,
 including brain scan data and tissue samples, available to other researchers.

More than 80 years ago, when it first opened its doors, the Montreal Neurological Institute offered up a bold vision for the future of neuroscience. Clinicians treating patients and researchers examining different neurological disorders and diseases began collaborating more closely than ever before.

The Neuro is about to shake things up again.

The institute recently announced that it will embark on a five-year project during which all of its research results, and all of the data associated with that research, will be made widely available. The Neuro won’t pursue patents on any of its discoveries.

“We’ve been doing a lousy job of advancing neuroscience,” says Neuro director Guy Rouleau. He isn’t talking about the institute he leads, but about the field in general. “We just aren’t progressing quickly enough. Part of the reason is that the brain is so incredibly complex. We need to find ways to do things differently.

“If 100 people look at the data that I work with, some of them will be interested in different things than I’m interested in,” says Rouleau. “It’s not that I don’t think I’m smart, but I have my own particular point of view. Other people looking at the same data would ask different questions than I would and that might lead to something new.”

“This takes the principle of open science to a new level,” says law professor Richard Gold, BSc'84, an expert on intellectual property and the biotechnology sector. Gold has been advising the Neuro as it mapped out its commitment to open science and he plans to study the ramifications of the institute’s five-year experiment.

“We have seen some significant commitments to open science before, but they were always directed towards specific outcomes. I don’t think we’ve ever seen an entire research institute do something like this,” says Gold. The best-known example of an open science project to date would be the Human Genome Project, which mapped out the entire human genome and made its results freely available.

A more recent example would be the Structural Genomics Consortium, an international multi-partner collaboration between university and industry scientists that hopes to further advance our understanding of the human genome while developing new drug treatments.

“As an institute that is largely funded by public money, I feel there is an obligation for us to share our work broadly,” says Rouleau. But figuring out how to do that sharing could pose some challenges.

“It isn’t enough to make the data available,” says Rouleau. “It needs to be made meaningfully available. We can’t just tell people to jump on a plane and come look through our books. My lab alone has almost one terabyte of data. We need to build the infrastructure that would make [sharing huge amounts of research data] possible. We have to develop an open science platform that doesn’t exist yet.”

The Neuro’s McConnell Brain Imaging Centre (BIC) has some expertise in that area. BIC is already something of a poster child for the principles of open science. It shares its research datasets with more than 30,000 registered users worldwide. It’s a principal player in such international collaborative initiatives as BigBrain, the first brain atlas offering microscopic detail. “They’ve already been drinking the Koolaid,” says Rouleau.

“This is a very exciting time,” says BIC acting director Sylvian Baillet of the Neuro’s commitment to open science. “It’s an extraordinary experiment.”

Baillet says that BIC “has a long tradition of data sharing,” but acknowledges that expanding that approach to the entire Neuro could be tricky. “With brain imaging, the data is digital. That’s easier to share than, say, tissue samples.”

Guy Rouleau is the director of the Montreal Neurological Institute.

Rouleau says that the commitment to open science was made after more than a year of consultations with Neuro researchers and staff. Town hall gatherings, meetings, surveys and retreats have all focused on whether the Neuro should take this step and, if so, what issues needed to be considered.

“We will not force people to be open,” stresses Rouleau. “That being said, the principle was unanimously approved by the Neuro’s scientists.”

Rouleau says that Neuro researchers could still pursue patents on their research results if they so choose, but the Neuro won’t fund or facilitate the process.

“Most of the research that goes on here is at a very early stage,” says Rouleau. Such research isn’t likely to have any immediate commercial potential. But new data about a particular brain mechanism might attract the attention of a pharmaceutical company that was already looking at those brain mechanisms. “Industry is very interested in this model,” says Rouleau. “They want to know the right direction to [point their own efforts at].”

“It takes time and costs money to patent something and most of these things die out,” says Baillet. “Very few breakthroughs end up being profitable, even for industry.” Taking patents out of the equation, says Baillet, “means one less bottleneck.”

Rouleau says that there will be one important exception to the Neuro’s open science experiment.

Advanced stage clinical trials designed to test the efficacy of promising drugs and treatments that have already been developed – and patented by the companies that developed them – will still go on.

“That’s beyond the limit of what we think should be open,” says Rouleau. “We want to help get medicines to the market. That’s good for our patients and we’re good at it. We want those trials to continue to take place here.”

The Neuro's plan is attracting a lot of attention. Both Science and Nature, two of the most biggest and most influential research journals in the world, have published news stories about the Neuro's commitment to open science. Companies have sent letters of support. "There has been a wave of positive feedback," says Baillet. "I think we're taking a lot of people by surprise."

"We think this is the right thing to do," says Rouleau. "We want to be able to prove it."

Gold will be involved in that process.

"We need to determine what success would look like," he says. "The first measure would be: Did we actually do what we set out to do? Did we make the data available? Did other people use it? We need to look at other things too. What impact will this have on philanthropy? Will industry help support this model? Will companies come to Montreal, sensing new opportunities?

"The Neuro is very much viewing this as an experiment," says Gold. "So, in that sense, even if it fails, it pays off. We'll know if this model works."