Allen Institute CEO Rui Costa on mapping the brain and paying early-career scientists decent salaries

Mapping the billions of cells that make up the brain is a task mammoth enough to keep hundreds of researchers across continents busy for years. But it’s just one of several ambitious projects the Allen Institute is helping tackle.

The nonprofit research institute, based in Seattle, was founded in 2003 by Microsoft co-founder and philanthropist Paul Allen to go after big and basic scientific questions. It’s now home to 827 employees, many of whom are researchers working in divisions dedicated to brain science, immunology, cell biology, and neuronal activity.

Allen Institute researchers recently played a leading role in identifying all the cell types present in the mouse brain, creating the first-ever complete map for a mammalian brain. That effort has laid the groundwork for scientists to now map the human brain, which is 1,000 times larger and home to an estimated 86 billion neurons plus just as many other cells that play important supportive roles.

The institute also recently announced a partnership with the Chan Zuckerberg Initiative and the University of Washington to create a synthetic biology hub to better understand changes in both healthy and diseased cells over time.

STAT caught up with Allen Institute CEO Rui Costa for a conversation about both of these projects, as well as about the institute’s emphasis on open, collaborative science and its efforts to support researchers at a time when there has been an unprecedented exodus of life scientists from academia.

This conversation has been edited for length and clarity.

There are a lot of nonprofit research institutes that do important basic research, such as Salk and Scripps. What makes the Allen Institute different?

We are complementary to those other institutes in the sense that the Allen Institute, as set up by Paul Allen, goes after big questions that are foundational, that go beyond one lab. For example, let’s map all the cell types of the mouse, macaque, and human brains. Let’s characterize immune cell types in teenagers, young adults, and seniors in health and disease. Let’s look into a cell to see what all its parts are doing in relation to each other as the cell is moving. These are moonshot projects, Manhattan-style projects with sometimes 70 people working together.

We share everything openly, not only the data but the tools, the standards, the algorithms, and, eventually, the discovery process. Three million people use our databases every year. We have 300 petabytes of data curated. That’s crazy. There are people right now finishing a Ph.D. without ever doing an experiment, just based on Allen Institute data. It’s almost a democratization of science.

Speaking of moonshots, researchers at the Allen Institute and elsewhere recently announced a complete map of the mouse brain, and you’re now working with others to map the human brain. Why is this effort important?

The first thing is it will enable understanding, in the sense that, at the moment, we don’t have a complete parts list of the mammalian brain. Having the complete parts list will not tell you how it works, but without having a list, I don’t believe we can understand the brain. That’s the basic idea, but I’ll tell you some aspirational goals. Right now, in neurodegeneration and dementia and especially in mental health, if you have a problem in the brain and you go to the hospital, you are treated very differently than if you have cancer. If you have cancer, they’ll tell you what type of cancer it is. They’ll tell you which cell types are involved. They will try to target that specifically. We’re not there where we can say with precision which brain circuits are affected and why. And I think we can get there.

Around the same time the mouse brain map was published, you announced a new synthetic biology research hub in partnership with CZI and UW. How will that work, and how will you know if the effort is a success?

It will work a bit like one of our moonshot projects. In the first two years, we’re going to try to create these synthetic cell recorders that are like smartwatches for cells and can sense what’s going on in the environment and write in the genome what happened. We’ll evaluate if it works and then will start getting into applications of diagnostics and therapeutics in immunology, and, eventually, the brain.

These are both big partnerships. I know it’s helpful to have large teams working on complex questions, but how do you keep everybody on the same page?

There are several things we do. One is that we have project managers who are well trained in science and who know what you need and who are constantly working to bring everyone together and organize meetings. It’s a significant effort, and it can be 20% to 25% of the work that goes into a project. We typically have weekly or even more frequent meetings with team members. And before you start a project, we agree to the standards of data collection.

The other thing that may help is that I think people see collaborating with the Allen Institute not as a competition where Allen’s going to get credit, but as a way they can get further and do things they can’t do alone. There’s room for everyone.

How is the institute organized? I’ve heard that you don’t have many or any graduate students and postdocs, which is unusual.

Each Allen Institute [division] is one entity, built around a moonshot project. There are leaders, maybe what you would call [principal investigators], but it’s not organized in labs and you can have different people leading interdisciplinary teams.

Everyone typically gets credit for participating, and if you produce a big dataset that you share with the world and that dataset is used widely, we count that as a success. You don’t need to have many first- or last-author papers. If you produced a new technology, a new dataset, if you made a discovery, you’re promoted here.

We have a few grad students in collaboration with [the University of Washington] and other universities. But we don’t have postdocs, we have what we call scientists. A Scientist 1 is the equivalent of a postdoc. The reason we call them scientists is that they’re already part of projects; it’s not only a training program. And we pay them slightly better than other institutions, maybe not as much as some industries. If they do well, they can be promoted to a Scientist 2, 3, senior scientist, etcetera. They could stay here 20 years if they’re doing well.

What’s the starting salary for a Scientist 1?

It’s around $100,000.

That’s a lot more than the average postdoc salary. We’ve been tracking the growing number of researchers leaving academia because of dissatisfaction with their wages and quality of life, among other issues. To what extent is that something you’ve been thinking about and trying to address?

Fantastic question. It’s not that money drives people; money drives people out. People cannot afford to stay in science. How can you do it if you have a family? What do you do? If you come from a disadvantaged background or ethnicity, this drives you out. If you want to be serious about changing how the next generation of STEM looks, you have to put your money where your mouth is.

These are real problems. Here at the Allen Institute, it was intentional for us to name our researchers as scientists, give them agency so they can lead some projects, and give them what we can afford as good salaries.

How are you able to do that?

First, we’re very lucky that we’re not for profit. Second, we have fewer people. You just can’t hire cheaper people; it’s not allowed here. We’re in a privileged position because we have some funds from [the Paul Allen estate], and that puts us in a different place. But we also get grants like others and other types of philanthropy. At the end of the day, it’s going to be a choice. Do you want to have a bigger workforce with lower benefits, or fewer people with higher benefits?