Advanced Deblurring of Electron Beam Induced Motion for High-Resolution CryoEM 3D Reconstructions using Electron Event Data
PI: Wen Jiang (Biochemistry and Molecular Biology)
RA support through the student’s primary department
Team members:
• Jean-Paul Armache, Assistant Professor of Biochemistry and Molecular Biology
• Sung Hyun(Joseph) Cho, Associate Research Professor; Co-directors of Huck Cryo-EM Facility
• Jenny Wang, Associate Research Professor; Co-directors of Huck Cryo-EM Facility
Departments and Units
• Eberly College of Science
• Huck Institute of the Life Sciences
Level of Effort
Two semesters at 10 hours a week (25%)
Connection to ICDS Mission
This project advances ICDS’s mission by leveraging data science and software development to support interdisciplinary research and advance biomedical imaging capability.
Engagement with ICDS
The PI Wen Jiang has engaged in discussions with ICDS director Guido Cervone about computational and storage support for cryo-EM. He has also met ICDS Assistant Director of Strategic Initiatives Gretta Kellogg about AlphaFold support and learned this funding opportunity. The team is committed to regular engagement through seminars and interdisciplinary collaborations facilitated by ICDS.
Project Description
1. Introduction and Background
Cryogenic electron microscopy (cryo-EM) has revolutionized structural biology, enabling the determination of macromolecular structures at near-atomic resolution. A critical challenge in cryo-EM imaging is the electron beam-induced motion of the sample. As the high-energy electron beam traverses the frozen specimen, inelastic scattering events deposit energy, leading to sample damage and nonuniform motion. This motion, particularly rapid movements during the initial phase of exposure, blurs the recorded images, thereby limiting the achievable resolution of 3D reconstructions.
In recent years, the development of fast direct electron detectors has partially addressed this issue. These detectors can output short movie sequences at tens of frames per second, allowing for computational motion correction to align the frames and reduce blurring. While current motion correction algorithms have significantly improved image quality, they still struggle to effectively compensate for the faster, more complex motions that occur, especially at the beginning of exposure. Our Falcon 4i camera offers a powerful capability known as electron event recording (EER). This mode captures the precise x/y coordinates and a high-resolution timestamp (at an internal recording frame rate of 320 frames/sec) for each electron hitting the sensor. This rich dataset provides an unprecedented opportunity to develop more sophisticated motion deblurring techniques.
2. Research Goal and Objectives
The primary goal of this project is to develop and implement a novel deblurring methodology for cryo-EM data that leverages the high temporal and spatial resolution of electron event recordings. This method aims to significantly improve the correction of electron beam-induced motion, especially the rapid initial movements, leading to higher quality 2D images and, consequently, more accurate and higher-resolution 3D reconstructions.
Specific objectives include:
• Develop a new data representation: Treat the EER data as a 3D point cloud (x, y, time) rather than a conventional stack of 2D image frames.
• Algorithm Development: Design and implement algorithms that can accurately track and model the complex, non-uniform motion trajectories of features within the sample directly from this 3D point cloud. This will likely involve techniques from image processing, point cloud processing, and machine learning.
• Deblurring and Reconstruction: Apply the derived motion information to generate motion-corrected images.
• Validation and Comparison: Evaluate the performance of the new method against existing state-of-the-art motion correction techniques using benchmark cryo-EM datasets and newly acquired data from the Falcon 4i.
3. Expertise and Expectations of the Junior Researcher
The junior researcher is expected to have expertise in image processing, machine learning, and programming skills in Python, numpy/scipy/scikit-learn, and GPU programming. He/she will play a key role in the algorithmic development, implementation, and validation phases of this project. Specific responsibilities will include:
• Developing Python-based code for processing EER data.
• Implementing and testing various image processing and machine learning algorithms for motion tracking and deblurring from 3D point cloud data.
• Benchmarking the new methods against existing approaches.
• Analyzing the results and contributing to the preparation of manuscripts for publication and presentations at scientific meetings.
• Collaborating closely with the PI and other team members.
4. Medium to Long-Term Goals
This project is expected to deliver:
• A novel open-source software tool for advanced motion deblurring in cryo-EM using electron event recording data.
• Demonstrably improved resolution and quality of cryo-EM images and 3D reconstructions, particularly for challenging samples exhibiting significant beam induced motion.
• New insights into the nature and dynamics of electron beam-induced motion at very high temporal resolutions.
• Peer-reviewed publication(s) reporting the algorithm, software, and applications.
The successful development of this method will have a significant impact on the field of structural biology. By enabling more accurate correction of beam-induced motion, it will push the boundaries of achievable resolution in cryo-EM, allowing for more detailed understanding of biological macromolecules and complexes. This could accelerate discoveries in fundamental biology and drug development.
5. Engagement and Mentorship
The junior researcher will be closely mentored by the PI, who has expertise in cryo-EM computational image processing and algorithm development. Regular meetings will be held to discuss progress, challenges, and future directions. The researcher will have opportunities to present their work within the research group and at institutional seminars. They will gain hands-on experience in cutting-edge computational methods applied to a high-impact biomedical problem.