Integrating gene expression and GO classification for PCA by preclustering
1 Institute for Molecules and Materials, Analytical Chemistry, Radboud University Nijmegen, Heyendaalseweg 135, 6525 AJ Nijmegen, The Netherlands
2 Department of Applied Biology, Faculty of Science, Radboud University Nijmegen, Heyendaalseweg 135, 6525 AJ Nijmegen, The Netherlands
3 MSD, Molenstraat 110, 5340 BH, Oss, The Netherlands
4 Centre for Molecular and Biomolecular Informatics, Nijmegen Centre for Molecular Life Sciences, Radboud University Nijmegen, Geert Grooteplein 28, 6525 GA, Nijmegen, The Netherlands
BMC Bioinformatics 2010, 11:158 doi:10.1186/1471-2105-11-158Published: 26 March 2010
Gene expression data can be analyzed by summarizing groups of individual gene expression profiles based on GO annotation information. The mean expression profile per group can then be used to identify interesting GO categories in relation to the experimental settings. However, the expression profiles present in GO classes are often heterogeneous, i.e., there are several different expression profiles within one class. As a result, important experimental findings can be obscured because the summarizing profile does not seem to be of interest. We propose to tackle this problem by finding homogeneous subclasses within GO categories: preclustering.
Two microarray datasets are analyzed. First, a selection of genes from a well-known Saccharomyces cerevisiae dataset is used. The GO class "cell wall organization and biogenesis" is shown as a specific example. After preclustering, this term can be associated with different phases in the cell cycle, where it could not be associated with a specific phase previously. Second, a dataset of differentiation of human Mesenchymal Stem Cells (MSC) into osteoblasts is used. For this dataset results are shown in which the GO term "skeletal development" is a specific example of a heterogeneous GO class for which better associations can be made after preclustering. The Intra Cluster Correlation (ICC), a measure of cluster tightness, is applied to identify relevant clusters.
We show that this method leads to an improved interpretability of results in Principal Component Analysis.