DTW-GO based microarray time series data analysis for gene-gene regulation prediction

Abstract

Microarray technology provides an opportunity for scientists to analyze thousands of gene expression profiles simultaneously. Due to the widely use of microarray technology, several research issues are discussed and analyzed such as missing value imputation or gene-gene regulation prediction. Microarray gene expression data often contain multiple missing expression values due to many reasons. Effective methods for missing value imputation in gene expression data are needed since many algorithms for gene analysis require a complete matrix of gene array values. In addition, selecting informative genes from microarray gene expression data is essential while performing data analysis on the large amount of data. To fit this need, a number of methods were proposed from various points of view. However, most existing methods have their limitations and disadvantages. To estimate similarity between gene pairs effectively, we propose a novel distance measurement based on the well-defined ontology structure for genes or proteins: the gene ontology (GO). GO is a definition and annotation for genes that describe the biological meanings of them. The structure of GO can be described as a directed acyclic graph (DAG), where each GO term is a node, and the relationships between each term pair are arcs. With GO annotations, we can hence acquire the relations for the genes involved in the experiment. The semantic similarity of two genes within biological aspect can be identified if we perform some quantitative assessments on the gene pairs with their GO annotations. In this chapter, we first provide the reader with fundamental knowledge about microarray technology in Section 1. A brief introduction for microarray experiments will be given. We then discuss and analyze essential research issues about microarray in Section 2. We also present a novel method based on k-nearest neighbor (KNN), dynamic time warping (DTW) and gene ontology (GO) for the analysis of microarray time series data in Section 3. With our approach, missing value imputation and gene regulation prediction can be achieved efficiently. Section 4 introduces a real microarray time-series dataset. Effectiveness of our method is shown with various experimental results in Section 5. A brief conclusion is made in Section 6. © 2011 Springer-Verlag Berlin Heidelberg.