摘要
以色谱工作站的积分数据为基础,定义了允许匹配峰组的域,提出了在域内根据色谱峰面积、保留时间计算各匹配峰组之间距离矩阵的公式,从而形成有向图。采用图论中的最短路径算法寻找可能的匹配峰组中的最佳匹配峰组。采用优化的匹配参数,对珍菊降压片、丹参提取物、柴胡皂苷提取物、人参皂苷提取物的HPLC谱图进行匹配并将有关结果与国家药典委员会推出的指纹图谱软件自动匹配结果作了比较。结果表明:本算法可最大程度地匹配可能的色谱峰并极少出现错配、漏配峰组,无需手动校正。
In order to match the corresponding chromatogram peaks in fingerprints, an original method, applying the shortest path algorithm in graph theory, was investigated and established. Basing on sample datapeak areas and their retention time that are obtained by chromatogram workstation, the potential matching domain of peak couples was defined, and the distance matrix formula in the domain was presented with these sample data as parameters, according to which directed graph could be gained. Finally, the best matching could be found by using the shortest path algorithm in graph theory. In the investigation, matching parameters were optimized：using the chromatogram of 40 batches Zhenju Jiangya tablets, and then they were used to match chromatogram peaks of Zhenju Jiangya tablets, and extracts of radix salvia miltiorrhiza, saikosaponin, and three samples of ginseng extracts prepared under different conditions. The results were compared with those obtained by chromatogram fingerprint matching software suggested by Chinese Pharmacopoeia Commission. The comparison indicates that the method can automatically match the potential corresponding peaks as well as possible, rarely mismatching and omitting peaks.
出处
《分析化学》
SCIE
EI
CAS
CSCD
北大核心
2006年第10期1454-1458,共5页
Chinese Journal of Analytical Chemistry
基金
"十五"攻关项目(No.2004BA721A45)
关键词
色谱指纹图谱
中药质量分析
谱峰匹配
有向图
最短路径
Chromatogram fingerprints, quality analysis of Chinese herb, matching of chromatogram peaks, directed graph, shortest path
