Flexible-body refinement for Cryogenic Electron Microscopy Applications

科学家们对生物分子的原子结构感兴趣，换句话说，分子是什么样子的。详细了解分子的样子为它可能的工作方式提供了重要的线索。如果我们能走得更远，在与其他生物分子或人工化合物(如药物)相互作用的过程中捕获分子，我们就能更清楚地了解它们是如何工作的。我们对生物分子结构的大部分知识来自X射线结晶学。然而，在过去的十年里，一种新的技术--电子显微镜(EM)变得流行起来。保存在液体溶剂薄膜中的单个分子被冷冻并放置在电子显微镜中，电子显微镜捕捉分子的图像。许多单独的视图可以组合在一起来构建分子结构的三维模型。在过去的几十年里，由于新类型的电子探测器和新的软件方法，电子显微镜得到了迅速的发展，导致了一场“分辨率革命”，使人们能够更好地了解分子。在最常见的情况下，分子的图像足够“模糊”，以至于我们看不到单个原子。因此，EM用户需要事先对分子结构或至少部分分子结构有一些了解。这种先验知识可能来自其他技术，如X射线结晶学或计算建模。然后，以前的模型可以被适配到EM图像中，以给出整个结构的指示，并允许理解大分子机器，如核糖体。以前的模型通常与真实结构不太匹配，要么是因为它来自不同的物种，要么是因为它被结晶扭曲了，或者是因为计算建模过程中的限制。因此，必须调整模型以适应观测到的电磁图像。这是使用Flex-EM这样的自动化软件来执行的，Flex-EM将结构分解成连续的较小片段并调整它们的位置以适应密度，以及使用3D图形进行耗时的手动建模。该项目的目标是采用一种现有的方法来扭曲一个3D图像以更好地适应另一个图像，并将其应用于从EM图像确定分子结构的几个问题。这种方法是考坦教授为解决X射线结晶学中的问题而开发的，但它具有足够的通用性，可以应用于其他问题。我们要解决的第一个问题是将已知的分子结构适配到3D EM图像中。移动场精化可以非常快速地确定模型的平滑变形，从而提高其与图像的拟合程度，而不是将模型拆分成各自单独拟合的较小片段。我们还将研究改善柔性分子的EM图像的问题。在这种情况下，3D EM图像是模糊的，因为它是由数千个粒子的2D图像组合而成的，每个粒子略有不同。我们将改进3D粒子图像，方法是将更相似的粒子的较小群集平均在一起，然后使用移动场细化来调整群集之间的差异，然后将它们平均在一起以生成最终的3D图像。该项目涉及为这些问题在现有计算机软件中添加新步骤，并以一种易于与现有软件集成的方式实施新方法。我们正在使用现有的软件工具，包括Flex-EM，而不是从头开始开发一套软件，以降低项目成本，提高成功的机会，并利用新的和现有方法的最佳功能。该项目产生的所有软件将通过现有的电子显微镜软件套件免费分发给学术用户。软件的源代码也将被分发，以便其他开发人员可以从中学习或修改。

Scientists are interested in the atomic structure of biological molecules, in other words what the molecules look like. Knowing in detail what a molecule looks like provides important clues to how it might work. If we can go further and capture molecules in the process of interacting with other biological molecules, or artificial compounds such as drugs, we get a clearer picture of how they work.Most of our knowledge of the structure of biological molecules comes from X-ray crystallography. However over the past decade a new technique, electron microscopy (EM) has become popular. Individual molecules held in a thin film of liquid solvent are frozen and placed in an electron microscope, which captures images of the molecules. Many individual views can be combined to construct a model of the structure of the molecule in 3 dimensions. Electron microscopy has developed rapidly over the past decades due to new kinds of electron detector and new software methods, leading to a 'resolution revolution' enabling a much greater understanding of the molecules.In the most common cases, images of molecules are 'fuzzy' enough that we can't see individual atoms. The EM user therefore needs to have some knowledge of the structure of the molecule, or at least parts of it, in advance. This prior knowledge may come from other techniques, such as X-ray crystallography or computational modelling. The prior models can then be fitted into the EM image to give an indication of the whole structure, and allowed large molecular machines such as the Ribosome to be understood.The prior model is generally only a poor match for the true structure, either because it came from a different species, or because it was distorted by crystallisation, or because of limitations in the computational modelling process. The model must therefore be adjusted in order to fit into the observed EM images. This is performed using both automated software such as Flex-EM which breaks the structure into successively smaller fragments and adjusts their positions to fit the density, and by time consuming manual modelling using 3D graphics.The aim of this project is to take an existing method called 'shift field refinement' for distorting one 3D image to better fit another, and apply it to several problems in the determination of molecular structures from EM images. The method was developed by Professor Cowtan for problems in X-ray crystallography, but is sufficiently general to apply to other problems. The first problem we will address is fitting a known molecular structure into a 3D EM image. Rather than breaking the model up into smaller fragments which are each fitted separately, shift field refinement can very rapidly determine smooth deformations of the model which improve its fit to the image.We will also look at the problem of improving EM images of flexible molecules. In this case, the 3D EM image is blurred because it is combined from 2D images of thousands of particles, with each particle being slightly different. We will improve the 3D particle image by averaging together smaller clusters of more similar particles, and then using shift field refinement to adjust for the differences between the clusters before averaging them together to produce final 3D images.The project involves adding new steps to existing computer software for these problems and implementing new methods in a way which can be easily integrated with the existing software. We are working with existing software tools, including Flex-EM, rather than developing a suite of software from scratch to reduce the cost of the project, improve its chances of success, and to exploit the best features of the new and existing methods.All of the software produced by the project will be distributed freely to academic users through existing software suites for electron microscopy. The source code for software will also be distributed so that other developers can learn from it or modify it.