By and large, academic science incentivizes researchers to “self-promote” in order to receive funding. In recent years, major journals like Science and Nature have written about how pressure on academic labs to publish breakthrough studies to thrive, affects the quality and reproducibility of published results.

Ten years ago, the Allen Institute in Seattle, Washington, received an initial $100 million seed grant to complete a single project: map where genes are expressed in the mouse brain. Unlike many other organizations designed with a singular goal, the Allen Institute has expanded beyond this initial project to become a key player in the neuroscience community. How? By applying an industry-style approach to scale up tried-and-tested scientific methods and by sharing their findings pre-publication. Not only does the Allen Institute provide benchmarks for the neuroscience community, it has also evolved to include its own unique brand of research laboratories and publishes work in academic journals.

HippoReads correspondent Cailey Bromer spoke with Dr. Carol Thompson, Scientific Program Manager at the Allen Institute, about the organizational structure behind this unique entity and her personal journey there.

CB: Can you tell me a little bit about how you came to the Allen Institute?

CT: I had just completed my Ph.D. in biochemistry at the University of North Carolina, Chapel Hill, where I studied non-visual circadian photoreception in mice. I was doing a lot of behavioral and molecular work toward this goal. After completing a short post-doc to wrap up some projects, I considered going into industry. I knew I was looking for something completely different. This was in 2004, right around the time the Allen Institute was beginning work on its inaugural project to map the expression of approximately 2,000 genes in the adult mouse brain using in situ hybridization, and I took a post-doc position there.

The Allen Institute got its start when Paul Allen donated $100 million for a three-year project to map the mouse brain. It was really important to the success of the organization that those of us working there at the time had a well-defined project. If we were working on a million things, the Allen Institute may have never gone anywhere. It was very exciting: everyone working there was focused on the atlas. By the time I arrived, the company had already grown to about 75 employees just in the first year.

We applied an industry mindset to the project: how could we scale things up, ensure quality control and reproducibility, and package our data as a product? We had to figure out how to go from methods (in situ hybridization) to a product.

CB: After the first three-year grant terminated for the initial project, how did the Allen Institute move forward from there? What role did you play in the transition?

CT: I started analyzing and interpreting the data we had collected in the first project, and I also researched how sleep deprivation affects gene expression in the mouse brain. From 2007, I transitioned into a new role and became the first “program manager.” My goal was to leverage all the tools, infrastructure, and processes we had built for the Allen Brain Atlas to map gene expression in the mouse across different points in development, to create the Allen Developmental Mouse Brain Atlas.The plan was to roll the developing mouse atlas as an interim project while we continued work on the Allen Human Brain Atlas, a longer-term project we were pursuing. The intended interim project was expected to take a nominal amount of time, because we thought we had the infrastructure in place. It ended up taking much longer than we expected; rather than being able to leverage our existing methods, we ended up having to tweak the in situ protocol for each of the seven developmental time points we had chosen to get good results. Instead of using one existing protocol, we had to develop seven additional ones.

CB: What are some things you consider to be essential for the organization’s success as you’ve watched the goals and structure of the Allen Institute evolve over time?

CT: Generating big data is foundational for Allen Institute to make its atlases. For this type of work, it’s important to have standardized protocols for methods. It’s easy to standardize data when we generate all of it in one place. There are other ways to achieve standardization by establishing community-wide standards and guidelines, but this is more difficult to maintain. We are able to provide benchmarks in established methods because of our standardized processes. The atlases were great: they were high-throughput projects that quickly went online. The whole point of the first generation of atlases (including the mouse connectivity and non-human primate atlases) was to generate data, ensure that the data was accurate and reproducible, and hope others would make use of it.

The Allen is also notable for its efforts to improve and facilitate open access and sharing data. We try to get the data we collect online before it’s published in academic journals. Otherwise, it could be two years before anyone could get access to this new information. We can afford to do this because we’re not solely funded by standard federal grants.

CB: You mentioned the Institute is working on a couple of different projects. How did the organization shift from its initial focus to accommodate multiple projects?

A couple years back, Christof Koch joined the Institute as the Chief Scientific Officer (CSO) and there was a big shift: he helped take the organization in new directions beyond the atlases. In addition to sharing our data and hoping it would fuel discovery within the larger community, we also wanted to make discoveries ourselves. We hired a lot of investigators: we weren’t just an industry anymore. We have an industry-style half, that does structured science, where we take scientific methods, scale them up, and make them robust, much like we did with the atlases. Then we also have the research-driven half, which, while similar to an academic environment, was also different.

CB: How so?

CT: When you work in an academic lab, your PI (principal investigator) basically lets you work on anything as long as you can find someone to pay for it. At Allen, you don’t have to apply for your own funding, though you are encouraged to do so. We want to avoid this thing where you’re only working for yourself because it’s limiting. We want people to be working toward something more impressive than what could be achieved in an individual lab which has to worry about publication record, grants and timeline. Without those constraints, we are able to do more collaborative, long-term projects. While we encourage investigators to be independent, their work is expected to fit into the larger goals of the organization. Their research is done in a way that complements the industry side; it helps determine which pipelines we’ll pursue next on the industry side, and sometimes even develops those pipelines.

CB: How is your role at Allen unique compared to careers you might have pursued in industry or academia?

CT: When I first came into the role, I was responsible for figuring out what it would look like. I talked to people in Seattle, in various businesses, and I learned a lot from pharmaceutical program managers. However, the “program manager” role in pharma is ultimately very different than the role we were developing, because pharmaceutical companies are profit-drive. Program managers at the Allen Institute, myself included, work towards scientific milestones and not monetary goals, which makes our targets more amorphous. We have to figure out how we corral the best teams to find and achieve our objectives. It’s a lot of fun to figure out. I see a trend toward this kind of management style even in academia, where more labs are involved in collaborative or multi-institutional projects. There is a growing appreciation for the need to map your way forward.

CB: Have you partnered with other companies or startups to support the Institute’s goals?

CT: Where we have needs, we do. For example, we worked with TissueVision, a small company in Cambridge, when we worked on the the mouse connectivity database. Their tools were used to section blocks of tissue and image them with a laser; this created a really nice 3D data set where you can see the projections we wanted to show.

CB: The Allen Institute started out with a seed grant. How is it funded today?

CT: We run as a non-profit organization, and we continue to be funded by Paul Allen, who provided our first grant. Today, our funding varies by project a bit more. We receive more pipeline research grants and are part of large consortium style grants: that’s how the Human Brain Atlas and the Mouse Spinal Cord Atlas were funded. We also have fundraising lists that go out. The Developing Human Atlas and the Non-Human Primate Atlas were funded through contract-style government grants.

CB: If you could give a seed grant for a single project today, like the one that helped the Allen Institute get started, what would it be?

CT: I think our current most important project is the one on cell-types, which was just launched. A lot of people have jumped on the “cell type project” and the BRAIN Initiative lists about five or seven awards around understanding cell types in the brain. We’re especially interested in understanding cell types in the cortex (the part of the brain associated with higher functions, including learning and memory). We’re using electrophysiology, morphology, and gene expression analysis to gather information on all the different cell types to understand how function relates to molecular properties. There are so many cell types that a large number of standardized samples is needed. In order to hit a huge goal like that, standardization is key.

CB: What do you see as the role for the Allen Institute in the future?

CT: Our mission is to fuel discovery by sharing data and we plan to continue doing that. There is a ton of data out there that is not for the public, but is owned by for-profit organizations. Pharmaceutical companies have amassed databases that would be useful for everyone. This isn’t to say that they aren’t important players in their own right, but these private databases don’t help the neuroscience community. We at the Allen Institute intend to continue being a focal point for open access data.

Image Credit: wikipedia

About The Author

Avatar photo
Academic Correspondent, Neuroscience

Cailey Bromer is a neuroscientist, writer, and lifetime learner, whose interests lie at the intersection of science, education, and culture. She holds a B.S. from Brown University and M.S. from University of California, San Diego, both in Neuroscience. She has dabbled with writing fiction and is a contributing writer of NeuWrite San Diego, a graduate student run writing group devoted to bringing Neuroscience topics to a lay audience.