CAREER: Cryo-EM Guided de novo Protein Fold Elucidation

本CAREER项目的主要目标是将教育和研究结合起来，开发一种新的蛋白质结构预测算法“EM- fold”，以提高低温电子显微镜（EM）数据中蛋白质结构解析的分辨率。EM-Fold概念的最终目标是解决Ad蛋白IIIa的结构。越来越多的大型大分子蛋白质复合物的结构研究通过低温电子显微镜产生了中等分辨率（5-10A）的密度图。目前还没有集成的方法可以将蛋白质结构从头构建到亚纳米EM密度图中。在该项目的目标1中，EM-Fold是通过推导二级结构元素的共识预测协议、开发基于知识的评分函数以及实现蒙特卡罗组装和优化算法来开发的。Aim II的重点是在蛋白质结构预测技术关键评估（CASP）实验期间测试EM-Fold，推导快速几何哈希算法以将结构模型拟合到低温电镜密度中，并使用模拟低温电镜数据集建立中等规模的实验基准。在目标III中，将记录一个分辨率提高的腺病毒冷冻电镜数据集，预测腺病毒蛋白IIIa的结构，并为蛋白IIIa结构分配详细的置信水平。在这个项目中进行的研究只是越来越多的计算结构生物学项目的一个例子，这些项目对学生教育提出了跨学科的要求。为了满足这一需求，我们设计了一种名为“教育研究之旅”的新颖综合教育理念，通过外展活动和课堂教育将学生招募到他们的第一次研究经历中。外展部分包括一年一度的分子生物物理学开放日，将100多名本科生带到范德比尔特大学，主要来自科学领域代表性不足的群体。开放日是一个专门的系列讲座的补充，“蛋白质折叠@家庭”，在其他本科院校和活动中呈现。“蛋白质结构预测”课程作为第二教学模块，为研究奠定了理论基础。课程材料和工作坊教程在网上提供，供其他教师和学生使用。在第三个模块中，本科生和高中生的研究项目与EM-Fold开发中的任务在适当的复杂程度上保持一致。

The primary goal of this CAREER project is to integrate education and research to develop a novel protein structure prediction algorithm"EM-Fold" that advances resolution in protein structure elucidation from cryo-electron microscopy (EM) data. The concept of EM-Fold is developed with the ultimate goal of solving the structure of Ad protein IIIa. An increasing number of structural studies of large macromolecular protein complexes by cryo-electron microscopy have resulted in intermediate-resolution (5-10A) density maps. Currently no integrated approach is available to de novo build protein structures into sub-nanometer EM density maps. In aim I of the project EM-Fold is developed by deriving a protocol for consensus prediction of secondary structure elements, the development of knowledge-based scoring functions, and the implementation of the Monte Carlo assembly and refinement algorithms. Aim II focuses on testing of EM-Fold during the Critical Assessment for Techniques of protein Structure Prediction (CASP) experiment, the derivation of a rapid geometric hashing algorithm to fit structural models into cryo-EM densities, and the set up of a medium-scale experimental benchmark with simulated cryo-EM data sets. In aim III a resolution-increased cryo-EM data set for adenovirus will be recorded, the structure of adenovirus protein IIIa will be predicted, and a detailed confidence level will be assigned to the protein IIIa structure.The research carried out in this project is but one example for an increasing number of computational structural biology projects that pose interdisciplinary requirements on student education. To address this need a novel, integrated educational concept called "Research in Education RidE" is designed as a step-wise journey recruiting students through outreach activities and classroom education into their first research experiences. The outreach component consists of an annual molecular biophysics open house that brings over 100 undergraduate students to Vanderbilt University, primarily from groups underrepresented in science. The open house is complemented by a dedicated lecture series, "Protein Folding @ Home," presented at other undergraduate institutions and events. As a second educational module, the course "Protein Structure Prediction" lays the theoretical foundation for research. Course materials and workshop tutorials are made available online for usage by other teachers and students. In the third module research projects for undergraduate and high school students are aligned with tasks in the development of EM-Fold at appropriate levels of complexity.