Collaborative Research: III: Medium: Systematic De Novo Identification of Macromolecular Complexes in Cryo-Electron Tomography Images

细胞内的大分子结构在控制生命的各种生化途径和细胞过程中起着关键的功能作用。识别细胞中新大分子的结构是我们理解生物分子如何共同维持生命和疾病如何发生的基础。低温电子断层扫描（cryo-ET）是一种革命性的成像方法，能够系统地识别和发现其原生细胞环境中的未知结构，具有近原子分辨率。然而，由于当前计算方法的限制，冷冻et的这一显著潜力仍未得到充分利用。现有的计算方法只能恢复已知的结构，或者需要大量的人工努力来识别未知的结构。为了填补这一空白，该项目旨在通过开发和整合一系列最先进的计算方法，开发一种计算管道，可以从冷冻et数据中自动识别新的和未知的生物分子结构。本项目开发的开源软件和算法将在生命科学领域有广泛的应用。所开发的计算算法将广泛应用于其他相关领域，如医学图像分析和通用计算机视觉。该项目将通过跨学科课程和直接参与卡内基梅隆大学和普渡大学的项目，培养不同背景的博士后、研究生和本科生。从项目中传播的知识将通过国内和国际在线计算生物学研讨会和黑客马拉松呈现给研究生、本科生和高中生和教师。Cryo-ET在单细胞中大分子复合物的结构可视化和空间定位方面具有独特的优势。本项目将开发三个关键技术，用于对Cryo-ET捕获的大分子进行从头结构鉴定：1)一种新的快速、穷举搜索的亚层析成像定位方法，用于改进大分子复合物从头结构发现。2)基于结构数据库的大分子复合物快速形状搜索新方法。3)然后，将所开发的方法集成到计算管道中，实现Cryo-ET图像中大分子复合物的自动大规模识别。该项目开发的管道和软件将被整合到公共数据库和开源平台AITom中，以便为结构和细胞生物学社区准备使用。总的来说，这个项目将把Cryo-ET提升到一个新的水平，它可以通过数据库搜索进行系统的大分子形状测定和鉴定。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Macromolecular structures inside a cell play key functional roles in various biochemical pathways and cellular processes that govern life. Identifying structures of novel macromolecules in a cell is fundamental for our understanding of how biomolecules work together to sustain life and how diseases occur. Cryo-electron tomography (cryo-ET) is a revolutionary imaging method that enables the systematic identification and discovery of unknown structures in their native cellular context with near-atomic resolution. However, this remarkable potential of cryo-ET is still unused due to the limitations of current computational methods. The existing computational methods can only recover known structures or require extensive manual efforts to identify unknown structures. To fill the gap, this project is aimed at developing a computational pipeline that can automatically identify novel and unknown biomolecular structures from cryo-ET data by developing and integrating a series of state-of-the-art computational methods. The open-source software and the algorithms to be developed in this project will have a wide range of applications in life science. The developed computational algorithms will be beneficial broadly in other related areas, such as medical image analysis and general computer vision. The project will train postdoctoral fellows, and graduate and undergraduate students of different backgrounds through interdisciplinary coursework and direct involvement with the project at Carnegie Mellon University and Purdue University. The knowledge disseminated from the project will be presented to graduate, undergraduate, and high school students and teachers through national and international online computational biology workshops and hackathons.Cryo-ET has a unique strength in visualizing structures and spatial localizations of macromolecular complexes in single cells. This project will develop three key techniques for de novo structural identification of macromolecules captured by Cryo-ET: 1) A novel fast and exhaustive search-based subtomogram alignment approach for improved de novo structural discovery of macromolecular complexes. 2) Novel approaches for fast shape search for discovered macromolecular complexes against a structural database. 3) Then, the developed approaches will be integrated into a computational pipeline that enables automatic large-scale identification of macromolecular complexes in Cryo-ET images. The pipeline and the software to be developed in this project will be integrated into the public database and the open-source platform AITom, so that they are ready to be used by the structural and cell biology community. Overall, this project will bring Cryo-ET to the next level which allows systematic macromolecule shape determination and identification through database searches.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.