Dave Van Veen: Data-Driven Cryo-ET Reconstruction to Understand Disease Mechanisms (Technical University of Munich)
Participating in a summer research internship at the Technical University of Munich (TUM) in Munich, Germany, offered a fertile environment for scholarly development. This ten-week stint marked the most extended period I’ve spent in Central Europe. Living in Germany granted me a lens into another country’s social landscapes, enabling me to draw parallels with my experiences in the United States. My time was spent collaborating with computational researchers who are improving imaging methods in cryo-electron tomography (cryo-ET), enabling us to understand cellular structures and mechanisms in a novel way. Engaging with German researchers during this period was an invaluable learning journey, deeply enhancing my academic perspective. I extend my sincere gratitude to the fellowship's benefactors for facilitating this enlightening opportunity.
While I was a guest researcher at TUM, I worked with my collaborators on improving imaging methods for cryo-ET. This is an exciting time to be working in the field for two reasons: (1) recent technological improvements in biological imaging have yielded a rapidly growing amount of three-dimensional cellular data of frozen cells. (2) the concurrent explosion of machine learning (ML) for “big data” image analyses provides opportunities for mining these images to understand structural mechanisms at subnanometer resolution. While there is great potential for cryo-ET imaging, it’s fundamentally limited by data corruptions in the acquisition process which distort cellular features and limit their resolution. We collaborated to mitigate these data corruptions using data-driven techniques, improving image reconstruction and hence enabling better insights from these images for structural biologists.
Spending ten weeks with this group in Munich allowed for a fruitful exchange of research ideas. Leading up to my internship, we’d each been developing algorithms to improve image reconstruction. Their focus was on creating the highest quality image with ML methods. Because these imaging volumes are so large, however, reconstruction methods are often slow and infeasible for use in a cryo-ET lab. As such, I directed my focus toward computational methods which improved the reconstruction speed. While concurrently leading separate efforts, being in the same lab space enabled us to combine our intuitions toward a larger, shared project of using ML methods to improve cryo-ET reconstruction more holistically. Furthermore, my experience with structural biologists at Stanford allowed me to help guide our project toward a contribution that will be valued by the biological community. While my collaboration with this lab at TUM is ongoing, we will submit our work for publication by the end of the calendar year.
After publishing this initial project, we plan to continue our collaboration to address a key factor which limits research progress on cryo-ET reconstruction. Because this problem requires a deep understanding of algorithms, physics, and biology, it’s difficult for any one research group to span these fields and make meaningful progress. This renders cryo-ET a daunting task for computational experts, despite the field’s immense potential for scientific impact. Hence our next collaborative project will be to establish tools and benchmarks for computational researchers to understand the physical and biological components of cryo-ET so they can build useful algorithms and subsequently validate them. We hope that bridging this gap between algorithms and practice will accelerate contributions by the broader research community in this exciting space.
Beyond lab research, spending the summer in Munich was an enchanting chapter of my life, underscored by the city's exuberant vitality under the generous summer sun. The city, steeped in a rich tapestry of history and architecture offered a fascinating insight into Bavarian culture. Each week included bike rides through Munich’s pulsating heart and its serene outskirts, unveiling a curious blend of tradition and modernity. The city's sprawling parks and natural havens became my sanctuary. A particular highlight was the invigorating swims with new friends in the Eisbach (“ice brook”), the cool waters whispering tales of age-old traditions and shared laughter under the azure sky. Journeying further, the Alps extended an irresistible invitation to majestic mountains and breathtaking vistas. The harmonious intersection of beautiful nature and Bavarian culture made for an idyllic summer.