UMass Chan Medical School researcher and Chan Zuckerberg Initiative Imaging Scientist Caterina Strambio De Castillia, PhD, and her team, including Alessandro Rigano, senior software engineer in molecular medicine, are on a mission to uphold one of the foundations of science: well-documented data, which in turn is essential for ensuring data quality, reproducibility and exchange. Dr. Strambio De Castillia, assistant professor of molecular medicine and a cell biologist, is the senior author on two papers and a co-author on four more published in the December issue of Nature Methods, dedicated to reporting and reproducibility in microscopy. The issue also includes a profile of Strambio De Castillia.
The publication focuses on advances made in recent years to improve the documentation, quality control, publication and sharing standards for bioimaging experiments. One of the central topics of the issue is the development of standards and tools for exchanging imaging results and metadata—information about the experimental, technical and image analysis context of image data.
Producing high-quality quantitative results and being able to replicate experiments are scientific cornerstones. And to ensure rigor and reproducibility, scientists need to agree on shared terminologies to describe details about how an experiment was performed, what the specifications were of technical equipment used and how the images were analyzed.
These aspects have sometimes gotten overlooked as scientific research has become more complex.
Strambio De Castillia explained that as scientists drilled further down into the molecular details of cellular functions over the past half century, it became important to know how the information gleaned could be integrated into a broader understanding of biological systems.
“We’re not just looking at all the DNA in the nucleus, all of the RNA, all of the proteins. We’re trying to figure out how all of these individual components work together, are dynamically organized in three dimensions and move over time, such as for example to ensure the proper functioning of the genetic material inside the nucleus,” she said. To achieve this goal, the quantitative and rigorous use of imaging data has become increasingly important in recent years.
To deal with this level of information complexity and learn how the informational content of image data could be managed and exchanged among researchers, Strambio De Castillia acquired an in-depth knowledge of computer science.
“I am really interested in how to organize information, so that we can talk with each other about experiments, we can ensure scientific rigor, and we can publish sufficient experimental and technical details to guarantee reproducibility and meaningful data exchange,” she said.
Building on her pioneering work on integrated imaging pipelines for viral particle tracking in collaboration with the Biomedical Imaging Group at UMass Chan, Strambio De Castillia was invited in 2018 to join the National Institutes of Health’s 4D Nucleome Consortium as an associate member to develop microscopy and image analysis metadata standards.
Metadata is “information about other data” that is needed to interpret results, Strambio De Castillia said. It provides the essential context to ensure results can be properly evaluated, interpreted and compared.
“Metadata exists in many different contexts, but in biology it started becoming something that people began paying attention to when we started doing genomics,” she said. “You just couldn’t keep track of everything you were doing, the materials and methods, in your notebook, the way you did before big data. This is a fundamental aspect of science but we kind of lost track of it a little bit when we started being flooded with data.”
Strambio De Castillia connected with a larger project called the Open Microscopy Environment, an international consortium that developed an initial model for how to organize this information.
With support from the Chan Zuckerberg Initiative, Strambio De Castilla’s productivity on building integrated image data management and quality control pipelines “skyrocketed,” she said.
As a result, she was invited to organize and lead the quality control and data management working group of Bioimaging North America (BINA). BINA is a consortium formed to bring together the bioimaging community in Canada, the United States and Mexico, foster communication, data sharing and technology dissemination among core facilities, promote training and professional recognition for imaging scientists and be a bioimaging advocate for institutional partners and funding agencies, with which to conduct strategic dialogue.
One of the key roles of BINA, now funded by the Chan Zuckerberg Initiative, is to serve as a counterpart to other global imaging organizations such as Global Bioimaging or the recently formed Quality Assurance and Reproducibility for Instruments & Images in Light Microscopy (QUAREP-LiMi) initiative.
As such, Strambio De Castillia is now involved at the forefront of a global effort conducted in the context of QUAREP-LiMi to develop standards and tools to improve reproducibility, comparability, downstream data analysis and re-use of imaging data; promote rigorous record-keeping, quality control and data management; develop international consensus on guidelines and tools to facilitate microscope assessment and calibration; and promote and disseminate guidelines and best practices.
Importantly, the novelty of this effort lies in the fact that QUAREP-LiMi was specifically designed as bottom-up consortium involving all members of the imaging community, including microscope users, custodians and manufactures, core facility managers, imaging scientists, publishers, standards organizations, and funders, Strambio De Castillia said.
“This creates a community that brings together all those that have a stake in ensuring that imaging can be used for scientific discovery with the specific goal of generating consensus solutions that can be adopted by all,” she said. “But it also presents a unified voice of what would be needed and what kind of investments need to be done at the institutional and funding level to really advance the field.”
Related stories on UMassMed News:
Caterina Strambio De Castillia named Imaging Scientist by Chan Zuckerberg Initiative
Chan Zuckerberg Initiative funds atomic-level imaging research in Grigorieff lab at UMMS
