论文部分内容阅读
运用二叉树生长及搜索技术对蛋白质及多个自旋偶合系统的样品体系的多量子滤波COSY及TOCSY谱峰进行划分,产生出对应于氨基酸结构单元的互相独立的自旋偶合系统集。划分的结果自动转化为自旋拓扑邻接表,便于蛋白质NMR谱峰归属过程中做进一步的图论处理。实验表明,对于混合的但互相独立的氨基酸样品体系,如果没有强偶合作用(个别含芳环的氨基酸分子除外),划分可获得良好的结果。对于蛋白质样品,谱峰覆盖情况较多,划分的结果不太理想,但对于蛋白质结构决定仍有重要意义,此时划分算法要与覆盖分辨技术相配合。
Binary tree growth and search techniques were used to partition the multiquantum filter COZY and TOCSY peaks in the sample system of proteins and multiple spin-coupled systems to generate independent spin-coupled system sets corresponding to the amino acid structural units. The partitioned result is automatically transformed into a spin-topological adjacency list to facilitate further graph theory processing during protein NMR peak assignments. Experiments have shown that good partitioning yields good results for mixed but independent amino acid sample systems without strong coupling (except for aromatic ring-containing amino acids). For protein samples, there are many peaks and peaks, and the result of partitioning is not ideal. However, it is still important for protein structure determination. In this case, the partitioning algorithm should be compatible with coverage resolution.