The study analyzes the relationship between the structure of organic compounds and their physical and chemical properties with cluster analysis method,uses clustering and hierarchical clustering techniques,conducts PCA reduction and standardization in the data pre-processing stage to ensure the reliability of analysis results,and divides the compounds into 4 groups by K-means clustering.It is found that there are significant differences in molecular weight,polarity,solubility and toxicity,etc.Grouped rough groups further reveal the similarity and stratification relationship between compounds.The results show that the properties of compounds in different clusters are obviously different.The compounds of cluster A have high solubility and low toxicity,which are suitable for the environment.Those of cluster B have high molecular weight and moderate toxicity,which are suitable for industrial field.Those of cluster C have high toxicity,and can be used for bioactivity studies.And those of cluster D have high solubility and stability for applications,which are suitable for the application under security requirement.It is proved that the cluster analysis is effective in understanding the structure-property relationship of organic compounds.This provides reference for compound classification,property prediction and application in medicine and materials science.
维普
万方数据