5 Things We Learned from the Seed Networks Kickoff Meeting
The Human Cell Atlas (HCA) is an international effort to generate a reference map of all the cell types in the human body. The Chan Zuckerberg Initiative (CZI) recently announced funding for 38 collaborative teams — CZI’s Seed Networks for a Human Cell Atlas — to support the continued development of the HCA. These groups bring together scientists, computational biologists, software engineers, and physicians. Seed Networks projects share goals to generate new tools, open source analysis methods, and significant data to help accelerate progress toward a first draft. Tools will be made available and data will be shared via the HCA Data Coordination Platform (DCP), a unified resource that will enable data sharing across researchers around the world.
At the end of July, we gathered this group for an inaugural meeting in Toronto, Canada. Over 200 people attended: more than 150 of the Seed Networks investigators, external partners from a variety of funders and institutions, and representatives from the HCA DCP, along with CZI program staff, computational biologists, and engineers. Building the HCA is a complex challenge that demands diverse contributions, and the goal of this first meeting was to help individual labs and projects understand the many activities across the HCA community.
The group covered a lot of ground in three short days, and we’re excited to share our top five learnings:
1. No one can build a Human Cell Atlas alone
The HCA community is growing and progressing rapidly. Efforts across the community are providing initial characterizations of a single tissue or organ system. However, there are very few researchers in the world who have the resources or expertise to dive deep, with statistical power and diverse methodologies, across every tissue in the human body. This is difficult even in a single tissue — first measuring diverse molecular features and how they differ in space; then analyzing these diverse methods and integrating them requires a lot of innovation across many domains. This effort will take many funders and scientists, and the Seed Networks kickoff meeting highlighted the importance of related efforts, such as the Helmsley Charitable Trust’s Gut Cell Atlas, to help move towards a first draft of the HCA. By focusing on our collaborative strengths, the community will be able to iteratively build toward a complete picture of the human body, with diverse and representative data types from tissues across the world.
2. Knowledge is power, and so are protocols
We learned from our first grant program that there is a need for better ways of capturing and sharing the process of science. In the past 18 months, we have seen increased use of protocols.io, a platform for sharing, discussing, and running protocols for a wide variety of methods and techniques. The HCA Methods Development Community group on protocols.io now has 100 publications and over 300 members, with protocols on many varieties of tissue dissociation, sequencing, and single cell analysis. We need to keep this momentum going, encouraging the community to share these resources early and often, even when the method is imperfect or still undergoing optimization.
It was clear from discussions on this topic that encouraging early submission of protocols and engaging the community to help refine them holds remarkable value for all involved. Early examples are protocols generated by Steve Henikoff and Samantha Morris, both of which now have improved versions, robust community commenting, and ever-increasing use. The willingness to share with the community early should be recognized and celebrated.
3. Fitting the puzzle pieces together isn’t easy, and computational biology is critical
Building a comprehensive reference for 30–40 trillion cells isn’t easy. Because of the growth of the community, pace of technology, and interest from the international community, the HCA has gone from conceivable to feasible — and part of this feasibility is better understanding some of the common bottlenecks that we face. As more data is produced, the need to harmonize and integrate this data continues to grow. The distributed nature of the HCA will require bringing together data across many axes: different platforms that analyze the same biomolecules, diverse but related measurements, diverse genetic backgrounds, differences over lifespan, tissue handling methods and many more. The role of computational biology in helping us understand these differences and coherently assembling a reference resource will be critical in the coming years.
4. Meetings kickstart the activation energy required to solve tough problems
It’s hard to dedicate time to a problem on your own. Meetings provide the space, time, and resources to get together with other experts, identify collective challenges, and ideate on solutions. We learned there is interest and demand for bringing together key members of different sub-communities to work through problems together. Some examples of possible future meetings include:
- Gathering technology and data cores. These individuals are absolutely critical to large initiatives like the HCA. They help standardize and ramp up technologies and promote consistency across major universities and centers.
- Bringing together students, postdocs, and staff scientists from different tissue communities. There are common challenges related to processing and sequencing challenges for specific organ systems.
- Convening computational biologists. This group is central to the HCA, and supporting discussion and solutions for common challenges as they emerge, along with benchmarking and promoting standards, will accelerate the dissemination of robust analysis tools.
5. Open science is hard in practice but critical to improving outcomes
Ed Fong, Scientific Director of The Neuro in Montreal, was the keynote speaker at the kickoff and spoke to the possibilities and challenges around open science. The HCA as a community looks to be progressive on this topic and is promoting early access to data and openness. Yet taking community principles and converting them to community action and norms is hard.
Dr. Fong’s talk highlighted some of these challenges, while showing results for the power of open science to attract great early career scientists and foster interactions with industry without defaulting to assuming everything is proprietary. CZI will continue to support this aspect of the HCA as it looks to pioneer a new model of distributed consortia and identify mechanisms to celebrate, evaluate, and reward contributions that are made early and openly.
In-person meetings have high value by bringing everyone together in the same place — an important aspect for this community where 45 percent of the projects are international collaborations. For this meeting, we wanted to provide time and space for investigators to gain a better understanding of the goals and deliverables for each project, and build a deep appreciation for projects working on similar questions. Our goal was to help networks form with a collective vision of cross-cutting topics so the community can begin to clarify a key question: What does a first draft of the HCA look like?
CZI is a major supporter of the HCA because of our belief in the community, technology, and the potential impact of this reference in advancing our understanding of the cellular mechanisms of disease. We need to investigate the cellular level to unravel the complexity of the human body and the many diseases that affect us and to find treatments and cures. We look forward to the next three years of learning with the Seed Networks investigators.
Authors
Jonah Cool, Science Program Officer
Fiona Griffin, Science Project Manager
Norbert Tavares, Science Program Manager
Arne Bakker, Director, Meetings and Community for Science
Andréa Clavijo, Science Event Planner
