Artificial Intelligence (AI)-enabled Cryogenic Electron Ptychography For Bio-macromolecule Imaging

用于生物大分子成像的人工智能 (AI) 低温电子叠层成像术

• 批准号: 2729815
• 金额: --
• 依托单位: University of Warwick
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2729815
• 关键词: Artificial; Intelligence; AI; enabled; Cryogenic

Cryogenic transmission electron microscopy (cryo-EM) with single-particle analysis (SPA) is a powerful method for visualizing a wide range of biological macromolecules in three dimensions (3D) at near-atomic resolution, which can provide direct insights into function and mechanism. Despite these revolutionary advances, it remains a challenge to deal with such small, heterogeneous and/or flexible molecules or complexes. An emerging strategy is based on cryogenic electron ptychography, which has been recently demonstrated by us for phase reconstruction of biological samples under low dose conditions. Ptychography is an emerging computational microscopy technique for acquiring images with resolution beyond the limits imposed by lenses, which has been applied to high resolution x-ray imaging. Due to its high phase-sensitivity, robustness to low electron dose data and the recovery of the sample wavefunction, electron ptychography represents a potentially disruptive change in the rapidly growing field of cryo-EM. The aim of this project is to provide a new computational image restoration framework for visualizing bio-macromolecules in 3D at near-atomic resolution. You will base upon artificial intelligence (AI) and machine learning techniques and develop a completely new computational scheme to dramatically speed up the time-consuming ptychographic reconstructions from a large quantity of electron diffraction, and recover high-fidelity phase-contrast images of biological macromolecules at high resolution. Combining with SPA, this AI-enhanced ptychography would potentially reveal and examine the 3D structures of the small molecules in an unbiased and comprehensive manner. You will be exposed to diverse and interdisciplinary research areas in applied mathematics, biological macromolecules, electron microscopy and cryogenics. You will experience an internationally collaborative environment, where you will closely collaborate with international-leading electron microscopists in the Rosalind Franklin Institute, a national research institute, dedicated to developing new technologies to tackle important health research challenges. You will have access to the state-of-the-art JEOL GrandARM with aberration correction, cryo stage, which is the very first and unique purpose-built cryo-EM in the UK for this category of work. You will also work with world-leading structural biologists in the Division of Structural Biology at the University of Oxford.

低温透射电子显微镜（cryo-EM）与单粒子分析（SPA）是一种强大的方法，用于以近原子分辨率三维（3D）可视化各种生物大分子，可以直接洞察功能和机制。尽管有这些革命性的进展，但处理这种小的、异质的和/或柔性的分子或复合物仍然是一个挑战。一种新兴的策略是基于低温电子重叠关联成像，这已被证明是由我们在低剂量条件下的生物样品的相位重建。重叠关联成像是一种新兴的计算显微技术，用于获取分辨率超过透镜限制的图像，已被应用于高分辨率X射线成像。由于其高相位灵敏度，低电子剂量数据的鲁棒性和样品波函数的恢复，电子重叠关联表示在快速增长的冷冻EM领域的潜在破坏性变化。该项目的目的是提供一种新的计算图像恢复框架，用于以近原子分辨率在3D中可视化生物大分子。您将基于人工智能（AI）和机器学习技术，开发一种全新的计算方案，以大大加快从大量电子衍射中进行耗时的重叠关联重建，并以高分辨率恢复生物大分子的高保真相衬图像。结合SPA，这种人工智能增强的重叠关联图将有可能以公正和全面的方式揭示和检查小分子的3D结构。您将接触到应用数学，生物大分子，电子显微镜和低温学的各种跨学科研究领域。您将体验到一个国际合作的环境，在那里您将与Rosalind富兰克林研究所的国际领先的电子显微镜密切合作，该研究所是一家国家研究机构，致力于开发新技术以应对重要的健康研究挑战。您将有机会使用最先进的JEOL GrandARM，具有像差校正，冷冻台，这是英国第一个也是唯一一个专门为这类工作建造的冷冻EM。您还将与牛津大学结构生物学部门的世界领先的结构生物学家合作。