An example of the use of the RIKEN gene expression database. (a) Topography plot of gut-endoderm-derived tissues from the RIKEN mouse dataset. The full dataset for these 14 tissues was filtered by total variance to produce a subset of 113 genes. This subset was clustered two-dimensionally using CLUSTER , and the resulting gene and sample order used to create the pictured topography plot in MatLab R12 (Mathworks, Natick, MA, USA). (b) Blade diagram of lung data from the RIKEN dataset. Asubset of 1,319 genes with the highest total variance was selected across the E16 lung, neonatal day 0 lung, and adult lung experiments. Genes with peaks at the edges of the diagram are enriched in the respective developmental stages compared to the adult sample. (c) Clustering of lung-expressed genes. The subset of genes described in (b) was clustered one-dimensionally with the whole embryo experiments using the order of samples shown. (i) A cluster containing surfactant protein A, a known lung-enriched gene; (ii) a cluster of ESTs enriched in the E16 lung. The RIKEN identification numbers and gene names (if known) are given.