Clustering algorithms fragment a dataset into various groups or clusters by distributing the objects in a particular group that have a high degree of similarity and the objects in different groups show more dissimilarity. Nearly 189 drugs which are known to act against thyroid disease with 33 significant property values are extracted from MalaCards database was selected as a dataset. Properties of drugs evaluated from Drug Bank database. In order to achieve better results in cluster analysis, Hopkins statistic was used to assess the clustering tendency of a dataset. A value of 0.3313 was obtained which suggests that the data is uniformly distributed and highly clusterable. Further, elbow and silhouette techniques and NbClust package employed to determine optimal clusters which resulted in 3 cluster solutions as optimal. With k=3, the k-means and k-medoids algorithm resulted in 3 clusters of various sizes, whereas hybrid k-means-k-medoids approach resulted in superior clusters sizes. Employing the hybrid approach reduced the number of negative silhouettes in analysis.\r\n\r\nKeyWords: k-means, k-medoids, hybrid, silhouette, R program, NbClust\r\n