Novel machine learning approaches for improving structural discrimination in cryo-electron tomography-Administrative Supplement

项目摘要 细胞冷冻电子断层扫描（Cryo-ET）使观察细胞的 细胞器和大分子复合物在纳米分辨率与天然构象。 然而，可获得的冷冻ET数据量的迅速增加带来了沿着一些主要的 我们将在本提案中及时解决分析方面的挑战。我们会设计新颖的数据- 驱动的机器学习算法，用于提高结构鉴别力和分辨率。在 具体而言，我们有以下几个具体目标：（1）我们将开发一种新型的自动编码器， 基于迭代区域匹配的倾斜序列图像无标记对齐算法 重建断层图像的分辨率提高;（2）我们将开发一个基于显着性的自动 挑选算法，用于更好地检测大分子复合物，并将其与联合收割机结合， 创新的2D到3D框架，以进一步提高结构检测精度;（3）我们将设计 用于子断层图像的姿态不变聚类的端到端卷积模型。这种模式的 产生将由新的子断层图像平均算法细化的初始聚类， 自动降低噪声和贡献小的子断层图像的权重;（4）我们将执行 通过使用先前报道的细菌分泌系统进行实验评价， 线粒体超微结构数据集，以提高最终分辨率。实现算法， 目标1-3，我们将开发一个用户友好的开源图形用户界面α-tom， 造福科学界。α-tom将与现有的软件进行系统的比较 包括IMOD、EMAN 2和Relion。以促进 分发，α-tom将集成到现有的软件平台Scipion和TomoMiner。我们 数据驱动的算法和软件不仅将促进和加速冷冻技术的未来使用， ET，而且可以很容易地用于分析现有的大量Cryo-ET数据， 提高我们对大分子结构、功能和空间组织的理解 复合物在原位。

Project Summary Cellular cryo-electron tomography (Cryo-ET) has made possible the observation of cellular organelles and macromolecular complexes at nanometer resolution with native conformations. The rapid increasing amount of Cryo-ET data available however brings along some major challenges for analysis which we will timely address in this proposal. We will design novel data- driven machine learning algorithms for improving structural discrimination and resolution. In particular, we have the following specific aims: (1) We will develop a novel Autoencoder and Iterative region Matching (AIM) algorithm for marker-free alignment of image tilt-series to reconstruct tomograms with improved resolution; (2) We will develop a saliency-based auto- picking algorithm for better detecting macromolecular complexes, and combine it with an innovative 2D-to-3D framework to further improve structure detection accuracy; (3) We will design an end-to-end convolutional model for pose-invariant clustering of subtomograms. This model will produce an initial clustering which will be refined by a new subtomogram averaging algorithm that automatically down weights subtomograms of noise and little contribution; (4) We will perform experimental evaluations by using previously reported bacterial secretion systems and mitochondrial ultrastructures datasets to improve the final resolution. Implementing algorithms in Aims 1-3, we will develop a user-friendly open-source graphical user interface α-tom to directly benefit the scientific community. α-tom will be systematically compared with existing software including IMOD, EMAN2, and Relion on simulated and benchmark datasets. To facilitate distribution, α-tom will be integrated into existing software platforms Scipion and TomoMiner. Our data-driven algorithms and software not only will facilitate and accelerate the future use of Cryo- ET, but also can be readily used on analyzing the existing large amounts of Cryo-ET data to improve our understanding of the structure, function, and spatial organization of macromolecular complexes in situ.