An Adaptive Segmentation Method Based on Gaussian Mixture Model (GMM) Clustering for DNA Microarray

Microarray allows us to efficiently analyse valuable gene expression data. In this paper we propose a effective methodology for analysis of microarrays. Earlier a new gridding algorithm is proposed to address all individual spots and to determine their borders. Then, a classical Gaussian Mixture Model (GMM) is used to analyse array spots more flexibly and adaptively. The Expectation Maximization (EM) algorithm is used to estimate GMM parameters by Maximum Likelihood (ML) approach. In this paper, we also addressing the problem of artifacts by detecting and compensate using GMM mixture components and artifacts data present in foreground and background spots are corrected by performing mathematical morphology and histogram analysis methods.