An oral history of an NSF Ideas Lab
About six months ago, a group of scientists traveled to a remote corner of Virginia to discuss olfaction. They were enticed--or warned, depending on your point of view--that the experience would be intense, the discussions would be collaborative, and there would be about $15 million in research funding on the table.
The event was titled "Cracking the Olfactory Code: A National Science Foundation (NSF) Ideas Lab." Held over five days at the Janelia Research Campus, it brought together a diverse group of about 30 scientists to accomplish a big task: create transformative research proposals that demystify olfaction.
Why focus on the sense of smell? Olfaction is an ancient system, evolutionarily speaking, and crucial for survival and propagation across the animal kingdom. Scientists have been researching it for more than a century, yet we still know little about how the brain processes and identifies odors.
Modern developments in systems, synaptic and molecular neuroscience--a deeper understanding of the brain in general--have put a breakthrough in reach. Olfaction can serve as a model system for studying neural coding, revealing insights about basic principles of neural activity and complex behaviors.
Ideas Labs are not new concepts. NSF has held others--including one to build an interactive Tree of Life and another on nitrogen-fixing in plants. They are based on a British model called Sandpit, which has been around since the early 2000s.
Ideas Labs are centered around a given, timely topic and designed to create a fertile, creative atmosphere--one that leads to fresh thinking and new approaches, and one that fosters team science. At NSF, they are a mechanism to accelerate innovative ideas, said Sridhar Raghavachari, a program director in NSF's Biological Sciences (BIO) Directorate, which funded the Ideas Lab, along with NSF's Mathematical and Physical Sciences Directorate.
"Ideas Labs supplement existing funding structures and give us the chance to support projects with the potential to transform research paradigms and solve intractable problems," Raghavachari said.
"Identifying ideas, especially ones that are collaborative--that's a gestation process," said Edda Thiels, another program director in the BIO Directorate involved in the Lab. "It's not like scientists don't get together; we're always getting together at conferences. You have collaborations, but they take years and years to develop. Ideas get lost; it's inefficient. But new interdisciplinary collaborations emerging from a five-day science camp or however you want to look at it--it really is unique.”
NSF issued a call for Ideas Lab participants in the spring; 30 individuals were selected from those applications by an NSF review panel of disciplinary experts. Those individuals formed five teams at the Ideas Lab; three of those teams were awarded funding in September after submitting a full research proposal that went through NSF's standard Merit Review Process.
The awards are part of NSF investments in the BRAIN Initiative, a coordinated research effort across multiple federal agencies and private partners to hasten development of new neurotechnologies to help researchers answer basic questions about how the brain works. Other recent NSF BRAIN investments include $10.8 million in cutting-edge projects to study brain circuits in action, and $12.8 million for neural and cognitive systems research.
What follows is a retelling, in their own words, from some of the participants of the Ideas Lab experience.
John Ngai, a professor of neurobiology at University of California, Berkeley, (Ngai was one of the Ideas Lab mentors--impartial researchers that NSF staff select to help guide the teams. Mentors essentially serve as real-time peer reviewers for participants.): Nobody'd done this before. Even though we were prepped, even though the facilitators--who had a lot of experience in doing these things--told us, 'Here's what it might look like' and 'Here's what you might expect,' the facilitators didn't tell us everything they had up their sleeves.
Marcelo Magnasco, a mathematical physicist at Rockefeller University: You input all this energy into [a traditional research] proposal, and you stand a 10 percent chance of getting through. It's pretty grueling. For the Ideas Lab model, you give up an entire week of your life. You go there, and they sequester you in some isolated place. But the chances of actually getting funded are much higher. I said, 'Well, I'll try anything once.'
Brian Smith, a behavioral neuroscientist at Arizona State University: They were, I think, deliberately a little bit vague, and that was good. If you come in a little bit unprepared, then you're a little bit more malleable.
The week began with structured, team-building activities. The initial emphasis on getting-to-know-you exercises frustrated some participants.
Katherine Nagel, a neuroscientist at New York University School of Medicine: I think the difficult part about it was that we sort of just had to surrender to the process for the initial part, the team-building part.
Don Katz, participant, a psychologist and behavioral neuroscientist at Brandeis University: The activities were a little but foofy, but they actually were getting at problems people have in science, which are: they're not willing to let themselves look stupid, and that inhibits them from actually looking really smart.
Lisa Stowers, a neuroscientist at The Scripps Research Institute: I was probably fairly skeptical once we started initially. Most of us feel that we're normally in a creative mindset. And I don't know that I was receptive to the idea of having someone further condition that mindset. But whatever they did, it worked.
Ngai: On the second day, on a Tuesday, we would have various sessions in the morning and the afternoon, and then the mentors would meet with facilitators and NSF program staff. I remember that late afternoon before dinner--speaking only for myself--we were all kind of seething. I said, you have a bunch of race horses in there, and you need to let them start running.
Katz: As someone who has studied psychology and behavior and has done clinical work, that background is solid. If you put 30 scientists in a room like that and you try and do it without those exercises? There might be some people who would tell you it's a waste of time. But let me tell you, not doing them would be an even bigger waste of time.
Magnasco: On the third day, they had two rounds of people in two circles of chairs, looking at each other. You were facing a random person in that meeting, and in three minutes you have to come up with an idea of what you would write a grant proposal on. Then everyone would stand up, rotate. I sat down with seven or eight people, of which six I had only met there at the meeting, I had to agree with them on what we would write a grant proposal on. That taught me that I could literally go with any person in the room and go write a grant proposal with anybody. I had something interesting to talk about with anybody.
Over the course of the week, the Ideas Lab facilitators (who were not NSF staff) relinquished more and more control, ending with the researchers forming teams and writing research proposals.
John Crimaldi, a fluid dynamicist at University of Colorado-Boulder: We had just gone out for a break, we had these frequent breaks, and Lucia Jacobs [Ideas Lab participant] came up to me and said, 'We need your help.' It was very organic like that, bumping into people in the hallway, 'Hey we have this idea. Do you want to maybe join our team?' It was a little tense because obviously you wanted to make sure you got on a team and there were definitely people who didn't for a while.
Stowers: For me, my first two groups were on topics that I didn't really have expertise or interest in. I knew within the first couple minutes of being in each group that it wasn't something I could stick with or offer anything to. It was sort of this frustrating and arbitrary process, and it sort of left me scrambling to find a group that I could offer some aspect of an experimental foundation. Wednesday night, I didn't really have anything. And Thursday, I had an idea for something, but I could only convince one other person that it was a good idea. So I had a team of two, which isn't a team. I was ready at the early part of Thursday morning to just be done with it: This was an interesting process, but it failed me. On a coffee break Thursday morning, I started chatting with someone and realized that I had a perspective that worked and that they liked.
Smith: They had this thing called Soapbox, where you could go up and write your name on a board so you could go and give a pitch for something. Literally get up on a soapbox. And you had five minutes. We did this, everybody sat down and listened, there's 30, 35 people in the room, and we got up and just gave this pitch. Two other people got up and said, 'We like this idea and we'd like to work with you.'
The teams worked through the night to present their research proposals on Friday morning.
Crimaldi: You could walk through the halls at four in the morning, and they have these glass-walled conference rooms. You couldn't hear what people were saying, but you could look in and see other teams furiously working on their presentations. It had this summer camp vibe to it, except there was $15 million in the other room.
Ngai: Typically, when we're asked to help on a program like this, it's as a reviewer. At the end of the day we did evaluate the proposals, but we also were deeply involved in helping to shape them. For many of us, being involved in a project from its most nascent stage through the final proposal was a unique and certainly novel experience. Even though the mentors didn't have our own skin in the game, there was still the sense, at least for me, that 'Oooh, we have to really work here.'
Crimaldi: We have been conditioned over the years, through traditional funding cycles--which this was not--to sort of know where the sweet spot is for risk versus reward. You need to propose something that is innovative and novel enough to get funded, but can't be too far out there. We're pretty calibrated to know where that spot is. The mentors, constantly--and I'm speaking only for my group--said we love what you're doing, but you need to go bigger. Do something that's transformative. And so they worked hard to push us beyond our precalibrated comfort zones, and they worked hard to make it clear to us that they weren't interested in funding a project that would be funded in a normal NSF funding cycle. They wanted something bigger and bolder and broader and riskier. It took us a couple of iterations being constantly told that. Eventually, I feel like we really did go big.
Katz: People kept hearing this word "transformative". Right now in neuroscience, transformative is interpreted to mean technically new. And everyone was sitting there thinking, 'What's a new technique?' and of course that's not the way to think of a new technique. My group tried to say we are going to change the way people think about olfaction. We couldn't amongst ourselves get a clean idea of the way people should think about olfaction.
Vijay Balasubramanian, a physicist at University of Pennsylvania: I was in a group that was trying to construct a story that had a very broad theoretical framework, that wanted to go all the way from the receptors to the cortex--basically the whole thing. It was really interesting because we were trying to do it really quickly and really well at the same time.
After the week ended, proposals were reviewed and certain teams were invited to submit full proposals.
Ngai: When I left there, I said never again. It was exhausting. It was mentally and emotionally exhausting. But maybe I'd do it again…
Magnasco: I really didn't get to read my email. We were busy constantly. For half an hour before dinner I would go take a shower in a rush, Skype my kids and read my email and that was it. We didn't materially have time. I don't think I got more than 12 hours of sleep for the whole five days. But I'm a father and I can survive these things.
Balasubramanian: We all felt like we had our hands on a completely new thing and that's really, really exciting to sort of work out what that means. And is it true? And do you believe it? And how do you test it?
Nagel: I think that it was very effective at getting people to talk to each other. You go to a meeting, everybody hangs out with their friends. The older people hang out with the older people, and the younger people hang out with the younger people.
Stowers: Being walked into this experience together I think really allowed me to both appreciate and begin to understand the voice and the role of these people that do research so differently from myself. I have my own jargon; they have their own jargon. We figured out a way to communicate what's going to be super valuable going forward.
Katz: I am in contact with two or three of the researchers from that meeting, and I think that collaborations will come out of that. It's like when you release a movie, and it doesn't make money. And you release it internationally, and it makes a little more money. And you release it on Amazon, and it's a success. This Ideas Lab will eventually pay out, not in what was funded there but in things that will happen over the next year or two and things that will come out of it.
Ngai: Science is done within a community of people, and I think people just expanded their communities.
Stowers: Right now, in this sort of crazy weird way, I'm super hopeful. If you look at my group's proposal, it seems like something that should have been solved a decade ago. The need for it is so obvious that it doesn't seem creative. But it hasn't been done yet because it is too big and too nutty for one lab to solve alone. Each of us in this collaboration will work on a little slice, and it will be very manageable. I can guarantee that it's definitely going to reveal all sorts of new things about the olfactory code in ways no individual lab could have achieved.