This paper, deals with a study of data mining techniques such as clustering, biclustering and triclustering. A large number of clustering approaches have been proposed for analysis of gene expression. However, the results of the application of standard clustering methods are limited. For this reason, concurrent clustering such as biclustering to find sub-matrices that are a subset of rows and a subset of columns from a two dimensional data set. Most of recently clustering of the 3D real dataset the triclustering techniques is implemented. Tricluster are constructed from two datasets by selecting a subset of features from each dataset and one shared subset of rows form amongst all the rows. This study reveals the journey of clustering to triclustering for gene expression data to identify the highest potential gene cluster or group.