“Not just a cog”: a Q&A on team science in cell biology

The Allen Institute for Cell Science was launched in 2014 with a lofty goal: to develop a holistic understanding of normal and diseased human cells—studying cells in their entirety, rather than one molecular complex or process at a time, as they transition from state to state. Thanks to the forethought of the Institute’s founder, the late Paul G. Allen, it is structured differently than a typical research nonprofit. The nature of the Institute’s goal requires a team science approach, where all 70 people work on projects that feed into that holistic understanding. This results in a level of large-scale, interdisciplinary work not possible in the typical PI-led laboratory.

As team directors at the Institute who both came from academia, Susanne Rafelski and Graham Johnson have a unique perspective on team science in cell biology. In an era of big data and large-scale research projects, more and more labs are turning to interdisciplinary teams to tackle tough questions in cell biology. Here Rafelski and Johnson answer questions about their experiences working with interdisciplinary teams. While scientists have to work within the constraints of their own organization and funding structure, Rafelski and Johnson think there are lessons from working in a team-science–based organization that could apply to any researcher who’s interested in thinking outside the walls of their own lab.

Rachel Tompa: What questions need a team science approach?
Graham Johnson: Some questions require reproducibility
and scale. For much of our work, the Institute conducts methodical, highly reproducible image collection using 3D live cell microscopy to produce open datasets that can address questions about variations in human cell structure and physiology, as well as changes between cell states. It requires meticulous, industrial caliber work to get at the subtleties in cells.

Susanne Rafelski: There’s work that, by its nature, requires a scope beyond that of an individual lab. For example, if the question requires data, experiment types, or specialized knowledge that are beyond the general interest or expertise of individually funded grants or projects, you might need a larger team to tackle that question. The kind of work we do, where we are looking at subtle variations in many different aspects of cells as they differentiate and transition to disease states, might not by itself be suited to an individual academic lab, since it requires the generation of cell lines, collection of large datasets, and new kinds of analysis and visualization methods.

GJ: Say you want to find out if a point mutation in a protein has an effect—there are many mutations where the cell doesn’t survive or exhibits some extreme morphology, and that’s the kind of research individual labs do very well. If the mutation makes a more subtle difference, and you need to see tens of thousands of cells to understand how that affects the entire system, that’s where large-scale team science can come in: to create and then help extract insight from the potential gold mine of information captured in this type of high-quality, high-replicate data.

RT: What are the challenges of team science?
SR: You don’t start a team science project with a full team in place. You have to find ways to adjust the structure of your team at different stages of maturity of the project, and you also have to adjust how the team functions as it grows, because a small team inherently functions differently from a large team and because individuals within the team grow and change.

GJ: In hiring for our teams early on, we came across hesitation among some candidates that they’d just become a cog in a machine. We all were trained in academia, and many of us came here directly from academia, where the emphasis is often on producing high-impact and numerous independent publications. There can be a fear that you won’t be recognized if you are not the first or last author on a high-profile publication. But we’ve found that our scientists and engineers are still sought after just as much for their individual knowledge, and for their exposure to cutting-edge approaches and team science, and they’re still asked to do just as many presentations, as they would have in an academic lab.

SR: It’s almost the opposite of being a cog in a machine if you do true teamwork. Everyone’s expertise is valuable and these big projects wouldn’t happen without each team member’s contributions. That feels really meaningful to our researchers.

RT: What’s the difference between collaboration and team science?
SR: There are a lot of different ways scientists can work together and many different flavors of team science. The main differences I see between the type of team science we practice at the Allen Institute and large consortia or smaller academic collaborations are that we give our full-time effort to our team projects, we are all under the same roof, and we are all dedicated to a pre-planned set of activities that fall under one overarching goal. It’s fundamental to our work that it’s well-planned, and everyone who works here accepts that major risk of putting all our eggs into one basket toward that holistic understanding. But the potential pay-off is huge.

GJ: In academia, collaborations often happen fortuitously, for example at scientific conferences. People meet who are on similar scientific paths but who might have different skillsets, and they see an opportunity and start working together. If you have the people who are all experts in their fields under the same roof working toward the same high-level goal, the nice part is that instead of meeting once a year at a conference, we cross paths nearly every day in the hallways or at the coffee machines, so those fortuitous interactions happen much more often.

RT: What do you like most about team science?
GJ: My career goal when I was in academia was the same as the mission of this Institute, to assemble data to understand how whole cells work. I guess I’m lucky to have found such a perfect match, but I find that I can do much more toward that specific goal here than I could have by continuing the trajectory of running my smaller academic lab, which required collaboration with data generating labs that typically had many other projects to worry about. I really like being able to fully participate in all aspects of the science. Whether it’s planning the big goals, contributing to discussions about methods of the data collection, or helping people across any of the seven sub-teams to develop tools for analysis, modeling, publication, and outreach—it’s great to see how details are coming together to begin addressing our grand questions.

RT: What has surprised you about working in this team science model?
SR: What surprised me is how fulfilling it is to work in such a dedicated team setting. In school, I hated it when they put you into arbitrary groups to do group work, especially when I ended up having to do extra work to compensate for other people. The reason I decided to give up my own lab, my little agile sailboat, to become part of this cruise ship is that it’s not a randomly selected group of people where some are going to wait for you to float them. It’s a carefully selected group of people who have all decided to be part of something bigger. That doesn’t always make it easier, but it makes it very powerful.