Project 2 Exploiting new sources of information in density-modification

项目概要/摘要 在密度修正和模型建立中开发新的信息源 这是一个开发算法的项目，用于构建代表分辨率低于3 - 3.5 A的大分子晶体结构的原子模型，结合密度修改和细化的能力，并使用Phenix环境中所有强大的工具来自动化结构测定过程。 该项目的成功将使结构确定更加稳健和快速，导致更准确的模型，并允许使用较低分辨率的数据进行结构确定。该项目将与伯克利、杜克和剑桥小组密切合作进行。所开发的方法的增强功能将允许在比目前可行的更低的分辨率下自动化结构测定，为世界各地的学术界，政府和工业界的结构生物学家节省大量的时间和金钱，并允许在手动进行结构测定时完全不切实际的错误分析和验证水平。 在凤凰项目的过去5年中，我们与伯克利、杜克和剑桥的小组合作，开发了全自动的程序，用于结构测定X射线晶体学，并使用MAD、SAD和MIR方法的任何组合。作为这些高层次程序的基础，我们已经开发了新的方法来评估实验电子密度图的质量，从密度图中识别非晶体对称性，多晶体平均，以及识别未知的 配体，以及一套用于快速建模的新工具。此外，我们还与剑桥小组和贝克实验室开展了一项令人兴奋的合作，将结构建模领域的联合收割机算法与晶体学领域的互补算法结合起来。 我们现在计划集中精力开发方法，以加速确定今天仍然具有挑战性的结构。这些具有挑战性的结构包括那些数据仅扩展到3.0 - 3.5 A或更差的分辨率的结构，那些具有弱实验相位信息的结构，以及那些具有来自遥远同源模型的分子置换信息的结构。为了实现这一目标，我们建议与伯克利，剑桥和杜克集团开发分辨率相关的模型构建工具，联合收割机基于晶体图案的模型构建与互补的能量为基础的建模工具，从结构建模领域，并开发一个框架，将合并密度修改和细化到一个更强大的程序。在这些新的努力的同时，我们将继续在开发用于大分子结构测定的集成、灵活和易于使用的Phenix系统方面发挥作用。 这项工作将使结构生物学家能够确定大分子的三维结构，这对理解疾病和维护人类健康至关重要。

Project Summary/Abstract Exploiting new sources of information in density-modification and model-bullding This is a project to develop algorithms for constructing atomic models representing macromolecular crystal structures at resolutions worse than 3 - 3.5 A, for combining the power of density modification and refinement, and for using all the powerful tools in the Phenix environment to automate the structure determination process. Success in this project will make structure determination more robust and more rapid, lead to more accurate models, and allow structure determination to be carried out with lower-resolution data. The project will be carried out in close collaboration with the Berkeley, Duke and Cambridge groups. The increased power of the methods developed will allow automation of structure determination at lower resolutions than is currently feasible, saving large amounts of time and money for structural biologists around the world in academia, government and industry, and allowing levels of error analysis and validation that are completely impractical when structure determination is done manually. During the past 5 years of the Phenix project we have worked with the Berkeley, Duke and Cambridge groups to develop fully automated procedures for structure determination X-ray crystallography with any combination of MAD, SAD, and MIR methods. As a foundation for these high-level procedures, we have developed new methods for evaluation of the quality of experimental electron density maps, for identification of noncrystallographic symmetry from density maps, for multi-crystal averaging, and for identification of unknown ligands, and a suite of new tools for rapid model-building. Further we have embarked on an exciting collaboration with Cambridge group and the Baker laboratory to combine algorithms from the structure modeling field with complementary algorithms from the crystallography field. We plan now to focus our efforts on developing methods that will accelerate the determination of structures that remain challenging today. These challenging structures include those with data extending only to resolutions of 3.0 - 3.5 A or worse, those with weak experimental phase information, and those with molecular replacement information from distant homology models. To achieve this goal, we propose to work with the Berkeley, Cambridge and Duke groups to develop resolution-dependent model-building tools, to combine crystallographic pattern-based model-building with complementary energy-based modeling tools from the structure-modeling field, and to develop a framework that will merge density modification and refinement into a single more powerful procedure. In parallel with these new efforts we will continue our role in the development of an integrated, flexible, and easy-to-use Phenix system for macromolecular structure determination. This work will allow structural biologists to determine 3-dimensional structures of macromolecules that are of key importance in understanding disease and maintaining human health.