Gene Expression Data Clustering Based on Local Similarity Combination

Abstract

Clustering is widely used in gene expression analysis, which helps to group genes with similar biological function together. The traditional clustering techniques are not suitable to be directly applied to gene expression time series data, because of the inhered properties of local regulation and time shift. In order to cope with the existing problems, the local similarity and time shift, we have developed a new similarity measurement technique called Local Similarity Combination in this paper. And at last, we’ll run our method on the real gene expression data and show that it works well.