Multidisciplinary Challenge (MDC): Parallel and Distributed Computing for Solving Large Structural Biology Problems

多学科挑战 (MDC)：解决大型结构生物学问题的并行和分布式计算

• 批准号: 9527131
• 负责人: Dan Marinescu
• 金额: $ 238.5万
• 依托单位: Purdue Research Foundation
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1995
• 资助国家: 美国
• 起止时间: 1995-09-15 至 2001-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9527131&HistoricalAwards=false
• 关键词: Multidisciplinary; Challenge; MDC; Parallel; Distributed

9527131 Marinescu Dr. Marinescu and his collaborators will develop algorithms and programs for real time data acquisition and analysis of crystallographic and cryEM data. These are required to process data produced at high rates. 10 Mbits per second or more, by new photon sources like the Advance Photon Source and by CCD detectors. Similar data rates are produced by state of the art electron microscopes equipped with CCD detectors such as the new Philips CM200 installed at Purdue. Real time analysis of such data streams require parallel and distributed computing. The research will be done by an interdisciplinary collaboration among members of Purdue's Center for Structural Studies and Department of Computer Sciences. The biological objectives are to provide improved tools for the structure determination of proteins, nucleic acids, and complexes such as viruses, virus-receptors, and membrane bound cell surface molecules using X-ray crystallography, cryo-electron microscopy (cyroRM), and 3D image reconstruction. An essential requirement for this is the fusion of structural biology and high performance computing. The computer science objective is to create new algorithms, parallelize old ones, and design methodologies for the high performance computing required in structural studies of biological macromolecules. %%% The goal is to provide an integrated solution for the computational side of structure determination from data acquisition to data analysis and data visualization and to incorporate into our system programs developed elsewhere, e.g., the CCP4 suite. This group is in a unique position due to the extensive experience in structural biology augmented by expertise in computer science and mathematics. This project builds on ongoing collaborative research between faculty in the Structural Biology group and Computer Science department. ***

9527131 Marinescu Marinescu博士和他的合作者将开发用于晶体学和cryEM数据的真实的时间数据采集和分析的算法和程序。 这些都需要处理以高速率产生的数据。 10兆比特每秒或更高，由新的光子源，如先进的光子源和CCD探测器。 类似的数据率是由最先进的电子显微镜配备CCD检测器，如新的飞利浦CM200安装在普渡大学。 这种数据流的真实的实时分析需要并行和分布式计算。 这项研究将由普渡大学结构研究中心和计算机科学系的成员进行跨学科合作。 生物学目标是提供改进的工具，用于使用X射线晶体学、冷冻电子显微镜（cyroRM）和3D图像重建来确定蛋白质、核酸和复合物（如病毒、病毒受体和膜结合细胞表面分子）的结构。 一个基本要求是结构生物学和高性能计算的融合。 计算机科学的目标是创建新的算法，并行化旧的，并为生物大分子结构研究所需的高性能计算设计方法。 我们的目标是为结构确定的计算方面提供一个集成的解决方案，从数据采集到数据分析和数据可视化，并将其纳入我们在其他地方开发的系统程序，例如，CCP4套件 这个小组是在一个独特的位置，由于在结构生物学的丰富经验，在计算机科学和数学的专业知识增强。 该项目建立在结构生物学组和计算机科学系教师之间正在进行的合作研究的基础上。 ***