Contamination removal methods in cDNA microarray data

Our objective is to detect and remedy the contaminated spots for cDNA microarray data. To check the existence of unusual spots, single linkage clustering is used to assess the background intensities. Then, K-means clustering method is applied to identify the contaminated area. We estimate the amount of contamination, for background and foreground, through the use of nonparametric spline regression and empirical cumulative distribution approach, separately. A simulation study shows that the performance of the recommended approach is promising.