Figure 1.

Identification and clustering of DEGs. A. DEGs expression profiles ordered by fold changes on day 5 in each model. The number in bracket represents the gene numbers including each model. Genes selected by both model (model 1 and 2) overlapped between model 1 and model 2. B. DEGs were classified into six clusters according to their temporal expression patterns using SOM clustering methods. The number in each panel represents the number of genes in each cluster. Normalized intensities are log2-values of signal intensity. C. The range of expression ratios of DEGs in each cluster was drawn with a box plot. The line in each box designates the median quartile (Q2). Dashed lines designate the threshold values (log2 ratio of +0.75 and -0.75) for determining clusters of genes whose expressions were changed significantly. The dotted boxes represent the clusters showing their inter-quartile ranges (IQRs) and are the outliers of the threshold, and the genes in those clusters were selected as gene sets for GO analysis.

Lee et al. BMC Genomics 2011 12:426   doi:10.1186/1471-2164-12-426
Download authors' original image