Table 1.
Statistics for reads and detected genes.
Figure 1.
Gene expression profiles across different stages of early zebrafish development.
(A) Distribution of RPKM values across the nine developmental stages. The vertical dashed line denotes the threshold above which the genes were determined as expressed. The log2-transformed RPKM values of genes at each stage were binned with interval size 1, and the number of genes falling within each bin was calculated and plotted as a function of log2-RPKM values. (B) Number of genes expressed, activated and inactivated in each developmental stage. Expressed genes at a stage were those with RPKM values above the threshold shown in A; activated genes were defined as those expressed at the present stage but were not expressed at the previous stage, whereas inactivated genes were those expressed at the previous stage but not expressed at the present stage. Stage abbreviations on the x-axis: 64 C, 64/128-cell; OS, oblong-sphere; 50 E, 50%-epiboly; 15 S, 15-somite; 36 h, 36 hpf, prim-25; 48 h, 48 hpf, long-pec; 60 h, 60 hpf, pec-fin; 72 h, 72 hpf, protruding-mouth; 1 w, 1-week, early larva.
Figure 2.
Transcriptome dynamics during early zebrafish development.
(A) The correlation of global gene expression profiles and number of differentially expressed genes between developmental stages. Gene expression correlations, as indicated by colors in the bottom triangular, were calculated as the Pearson correlation coefficients using the RPKM values of all the detected genes. Differentially expressed genes were defined as genes with statistically significantly different raw read counts (FDR<0.001) and > = 2-fold changes in the RPKM values between two stages, and their numbers are represented by the colors in the upper triangular. Color bars at the bottom and the right side indicate the scale for correlation coefficients and the numbers of differentially expressed genes, respectively. The two squares with bold borders indicate the correlation and number of differentially expressed genes between the two replicates for the 60 hpf stage. (B) Number of genes significantly up-regulated or down-regulated during developmental stage-transitions. (C) Principal component analysis (PCA) about developmentally regulated genes. The graph illustrates the distribution of each development stage in the space of the first three principal components (PCs). (D) The loading coefficients on the first three PCs for each developmental stage. Stage abbreviations on the x-axes in (B) and (D) are as in Figure 1B.
Figure 3.
Distinct gene expression patterns in early developing zebrafish.
(A) Component planes obtained by self-organization map (SOM) analysis about developmentally regulated genes. Each component plane, except for the one marked as ‘clusters’, illustrates the normalized gene expression for one stage, with colors varying from red via yellow to blue indicating high to intermediate and low expressions. The ‘clusters’ plane shows the clustering of the units on the SOM planes. Map units at corresponding positions in each plane represent the same set of genes, with the number varying from several to hundreds. Clustered regions on the SOM plane represent more coherently expressed genes. Regions are numbered and distinguished by different colors. (B) Line-plots showing gene expression patterns for each identified cluster in A. The expression values are log2-transformed RPKM values. In each sub-graph, marks on the up-left corner denote the respective clusters on the ‘clusters’ plane in A and the grey line demonstrates the expression threshold. Stage abbreviations on the x-axis are as in Figure 1B.
Figure 4.
Genes preferentially expressed in one developmental stage of zebrafish.
The heatmap at the left side visualizes the self-division of preferentially expressed genes of each stage. A gene was taken as preferentially expressed at one stage when it has both significantly higher numbers of detecting reads (FDR<0.001) and higher RPKM values (> = 2-fold) at this stage compared with that at any other stages. In the graph, color gradient illustrates the z-scores of the expression values of the genes, which were calculated as the mean-centered RPKM values divided by the standard deviation for each gene, separately. Right to the heatmap shown the number of preferential genes and selected significantly enriched GO terms (Fisher’s exact test P<0.05, see texts for details) for the corresponding stages indicated by the grey lines. Numbers in parentheses represent genes annotated to the corresponding GO terms in the preferentially expressed gene group and the genome, respectively. Stage abbreviations below the heatmap are as in Figure 1B.
Figure 5.
The distinct expression patterns of transcription factors in early developing zebrafish.
(A) The expression patterns of transcription factor (TF) families expressed during early development of zebrafish. Shown are the log2-transformed median expression values for each TF family. The TF families are clustered into three groups, each with distinct expression patterns, as marked by the Roman numbers. (B) Comparison of the clusters of TF families and the clusters of developmentally regulated genes. Labels at the x-axis denote clusters identified by the SOM analysis in Figure 3. (C) The expression patterns of the MBD gene family. (D) The expression patterns of the Sox genes. Stage abbreviations on the x-axes are as in Figure 1B.
Figure 6.
The expression patterns of the Wnt signaling pathway and the cyclin genes.
(A) and (B) show the log2-transformed expression values of the Wnt gene family and the Wnt receptor genes, respectively. (C) The similarity and dissimilarity in the expression patterns between the Wnt genes and Wnt receptors. Color gradient represents Pearson’ correlation coefficients of the expression values. (D) The log2-transformed expression values of the genes encoding cyclines. Stage abbreviations on the x-axes are as in Figure 1B.