Krysiński J
Lehrstuhl für Arzneimitteltechnologie, Medizinischen Akademie, Poznań, Polen.
Arch Pharm (Weinheim). 1994 Apr;327(4):247-54. doi: 10.1002/ardp.19943270410.
Relationships between chemical structure and antifungal activity of 72 quaternary pyridinium chlorides were analysed using the method of rough sets. In the information system the compounds are described by eight condition attributes and divided into three classes of activity. Using the rough sets approach a smallest set of four condition attributes significant for a high quality of classification and accuracy of classes has been found. The resulting decision algorithm describes relations between structure and antifungal activity in terms of significant condition attributes. It may be helpful in predicting structures of new antifungal compounds to be synthesized.
采用粗糙集方法分析了72种氯化季铵吡啶的化学结构与抗真菌活性之间的关系。在信息系统中，这些化合物由八个条件属性描述，并分为三类活性。使用粗糙集方法，已找到一组最小的四个条件属性，这些属性对于高质量的分类和类别的准确性具有重要意义。所得的决策算法根据重要的条件属性描述了结构与抗真菌活性之间的关系。它可能有助于预测待合成的新型抗真菌化合物的结构。
