论文部分内容阅读
本文介绍了一种利用计算机从C-13NMR谱数据自动推断简单化合物结构的方法.该方法基于子结构/部分光谱拓扑相关模型从未知物分子式及C-13NMR数据推导出候选子结构以及对拼装这些子结构的约束条件,然后在这些约束条件之下利用部分重叠的方法将这些子结构拼装成候选结构.解析过程中每一步推导到的子结构或结构都经过评估,以便及时删除与已知数据相矛盾的(子)结构.评估的主要方法是模拟候选(子)结构的(部分)C-13NMR谱并与已知数据进行比较.解析过程是自动进行的,同时使用者可以根据自己的经验和直觉对推理过程加以干预以提高解析效率.对文献上报道的247个单官能团化合物作为未知物进行解析,成功率为97.2％.
This paper presents a method to automatically deduce the structure of simple compounds from C-13 NMR spectra using a computer based on the sub-structure / partial spectral topological correlation model to derive candidate substructures from unknowns molecular formulas and C-13 NMR data and to assemble these Sub-structure constraints, and then under these constraints using partial overlap method to sub-structure these substructures into candidate structures.Against the sub-structure or structure derived from each step of the analysis process have been evaluated in order to promptly delete the data with known data Contradictory (subs) structure. The main method of assessment is to simulate (partial) C-13 NMR spectra of candidate (subs) structures and compare them with known data. The parsing process is automated and the user can, based on his own experience And intuition to intervene the reasoning process to improve the efficiency of the analysis.According to the literature, 247 monofunctional compounds were analyzed as unknowns, the success rate was 97.2%.