Geometry and Bayesian statistics to reconstruct protein radical structures from ENDOR spectroscopy (A01)

利用几何和贝叶斯统计从 ENDOR 光谱重建蛋白质自由基结构 (A01)

• 批准号: 456827906
• 金额: --
• 依托单位: Institut für Mathematische Stochastik (IMS)
• 依托单位国家: 德国
• 项目类别: Collaborative Research Centres
• 资助国家: 德国
• 项目状态: 未结题
• 来源: https://gepris.dfg.de/gepris/projekt/456827906?language=en
• 关键词: Geometry; Bayesian; statistics; reconstruct; protein

We reconstruct 3D structures of protein radicals, by developing new methods of electron-nuclear double resonance (ENDOR) spectroscopy and solving the statistical inverse problem which leads from measured spectra to geometric and electronic structure. This involves devising a realistic error model taking into account, among others, spin dynamics effects and new methods of Bayesian statistics on manifold, e.g, fast Markov-Chain-Monte-Carlo algorithms and conditioning molecular dynamics on observed spectra. As a representative biological system, we study the ribonucleotide reductase (RNR) from E. coli, a prototype of all radical enzymes and fundamental in DNA replication and repair.

我们通过开发新的电子-核双共振（ENDOR）光谱方法和解决从测量光谱到几何和电子结构的统计逆问题，重建了蛋白质自由基的三维结构。这包括设计一个现实的误差模型，其中考虑到自旋动力学效应和贝叶斯统计在流形上的新方法，例如快速马尔可夫链-蒙特卡罗算法和根据观察光谱调节分子动力学。作为一个具有代表性的生物系统，我们研究了大肠杆菌中的核糖核苷酸还原酶（RNR），它是所有自由基酶的原型，是DNA复制和修复的基础。