Missing value estimation in microarray data using coregulation and similarity of genes

Microarray experiments usually generate multiple missing values in gene expression data sets due to several reasons. In the paper, a robust method has been proposed to estimate the missing values in microarray data using biological knowledge of the genes. Missing values are imputed based on the similarity in their characteristics patterns observed among the co-regulated genes. In this approach, first the microarray data are normalized and then the normalized data is discretized to measure similarity between the genes. The estimation accuracy of the proposed method is compared with the existing K-Nearest Neighbor based method and Pattern Similarity Matching (PSM) considering 4000 genes generated from 192 experiments. The experimental results exhibit that the proposed method outperforms other methods in terms of accuracy.