In 2022, the Mark and Lisa Schwartz AI/ML Initiative was launched at the Ragon Institute to explore the boundaries of artificial intelligence and machine learning within the context of immunology. Its first major project was the development of the MUNIS epitope predictor by the Gaiha Lab in collaboration with MIT. 

Now, Ragon faculty member Sophia Liu, PhD, is using it to develop a computing infrastructure to support the 26 labs at the Ragon Institute.

“The inspiration for this project came directly from our researchers’ needs—so many labs have faced barriers to accessing computational resources, and this initiative addresses that challenge directly,” Liu said.

Computational resources are the backbone behind much of the work done in any lab; whether sequencing data, analyzing complex datasets, or identifying cells to be targeted in vaccine designs, computation is a resource just as valuable as microscopes or autoclaves.

Currently, the Ragon community relies on distinct resources from our member institutions of Mass General Brigham, MIT, and Harvard. However, the fragmented nature of these resources means that the approach of each researcher here at the Ragon—whether for data processing or resource requests—can differ significantly, making it difficult to compare results, share data, or collaborate effectively.

“It’s great that we are all part of different institutions, but we need something that unites us,” Liu explained. “Without a unified system, it’s very hard to share ideas and collaborate efficiently.”

Liu’s vision, funded by the Schwartz Initiative, is to create a fully integrated computational infrastructure accessible to all labs at the Ragon Institute. Collaborating closely with the Ragon IT team, Liu will oversee the development of this infrastructure, beginning with the procurement of specific GPUs and CPUs to build the foundation. This resource will integrate existing tools and resources into a unified framework, simplifying access and usability for all researchers at the Ragon Institute, and serving as a basis for developing future methods.

“We hope to adapt tools like protein structure prediction to accelerate antibody-based research,” Liu said.

Currently, Liu and the IT Team aim to launch the platform within three months and begin a one-year pilot program. A big part of that will be gathering feedback from the community as the infrastructure is tested and applied by the labs.

Monthly computational meetings will provide a forum for knowledge exchange, community feedback, and iterative improvements to the infrastructure with the goal of fostering a computationally-engaged research community at the Ragon.

“In the short term, our focus is community-building—creating a space where researchers can explore computational methods and share insights and pipelines efficiently,” Liu explained.

Looking towards the future, this project has the potential to profoundly influence immunology research long-term at the Ragon, enabling ambitious computational projects across the institute.

“Looking ahead, we envision using this computational infrastructure to drive larger-scale projects that tackle significant questions in immunology, ultimately accelerating scientific discovery,” said Liu.

The Schwartz family’s critical role as funders of this impactful initiative, as well as their support for integrating computational innovation into immunological research, is invaluable to the Ragon Institute’s transformative research. 

“The community has often expressed that we need something like this,” Liu said, “and now we’re achieving that dream that our institute can take computational techniques and accelerate our discoveries in immunology.”