Figure 2

Clustered heat maps. Clustered heat maps for the seven host gene clusters (H) and the corresponding target gene expression profiles (T). For all three time course datasets only clusters with more than five host genes are shown. Each row corresponds to one gene expression pattern, each column to a measurement. Time dependent measurements are shown in ascending order from left to right. The expression level of each gene is standardized so that the mean is set to 0 and the standard deviation is 1. Expression levels above and below 0 are color-coded: red indicates for high and green for low expression levels, respectively; black for zero expression values. Biological replicates of the three datasets are in order from Rep. 1 to Rep. 2 and Rep. 3, respectively. Colored subtrees in the dendrogramm derived from hierarchical clustering denote for co-expressed (green) or anti-correlated (red) gene expression of predicted targets. (Somitogenesis) The dataset splits up into three host gene clusters, SG I with 13, SG II with 21, and SG III with 7 host genes. (Neurite Outgrowth) Two cluster with 10 (NO I) and 17 (NO II) host genes could be identified with similar behavior of host and target genes in both replicates. (Stem Cell Development) Two host gene clusters containing 9 (SCD I) and 8 (SCD II) host genes were identified. All host and target genes show similar behavior in all three replicates. For each dataset, flipped expression patterns between the host/target clusters are striking (SG I vs. SG II; NO I vs. NO II; SCD I vs. SCD III).