Fig 1.
(A) Average log2 FPKM values over all genes at the 12 ages. (B) Standard deviation of the log2 FPKM values at the 12 ages. (C) Total number of significantly expressed genes among the 7,289 significantly differentially expressed genes at the 12 ages. (D) Distribution of the gene expression rank (GER) of the 7,289 significantly differently expressed genes, describing the relative ranking of gene expression at the 12 ages. (E) The median gene expression rank (1–12) for each age. (F) The standard deviation of the gene expression rank (1–12) for each age.
Fig 2.
The average expression pattern of the different protein types over the 12 ages.
Fig 3.
Selected examples of differential expression of known alternative splicing transcripts.
Fig 4.
Fundamental gene expression patterns in mouse liver development.
(A.1) Plot of the factor loadings from the exploratory factor analysis of the significantly differentially expressed genes at the 12 ages. Higher loadings indicate days with gene expression that is highly correlated with the factor. The patterns of the factor loading graphs describe the relative contribution of the factors to the observed gene expression over time. (A.2) The average temporal expression patterns of genes in the six clusters. (B.1) Expression heatmap of hierarchically clustered genes correlating with factor 1. (B.2) Standardized (zero mean and unit variance) gene expression pattern of genes positively correlated with factor 1. (B.3) Standardized gene expression pattern of the genes negatively correlated with factor 1. (C.1) Expression heatmap of hierarchically clustered genes correlating with factor 2. (C.2) Standardized (zero mean and unit variance) gene expression pattern of genes positively correlated with factor 2. (C.3) Standardized gene expression pattern of the genes negatively correlated with factor 2. (D.1) Expression heatmap of hierarchically clustered genes correlating with factor 3. (D.2) Standardized (zero mean and unit variance) gene expression pattern of genes positively correlated with factor 3. (D.3) Standardized gene expression pattern of the genes negatively correlated with factor 3. The perforated lines plot average expression. The color-bar represents the relative log fold-change of genes relative to the average expression over the 12 ages.
Fig 5.
Functional analysis of gene clusters.
The activation or suppression status of different biological functions associated with genes in (A) Perinatal and Neonatal, (B) Prenatal and Adult, (C) Neonatal, (D) Neonatal and Adolescent, (E) Adolescent and Adult, and (F) Adult groups respectively. Red represents activation and blue represents repression of a biological function. The individual biological functions represented in the labeled sub-clusters are listed in S7 to S11 Tables.
Fig 6.
(A) Heatmap representing the number of genes targeted by each of the upstream regulators in the six temporal groups. Upstream regulators (rows) are ordered according to the hierarchical clustering (distance measure: correlation, linkage function: average) of the hit count matrix of the number of target genes. The red intensity is proportional to the number of targets. (B) Heatmap showing the temporal expression patterns of the hierarchically clustered upstream regulators (distance measure: Euclidean, linkage function: Ward). (C) Bar graph showing the hypergeometric p-value of the significance of association of upstream regulators in each sub-cluster in B with genes in each of the six temporal groups.
Table 1.
Overlap of upstream regulators between developmental periods.
Fig 7.
Top 15 upstream regulators with the highest number of target genes.
The top 15 upstream regulators with the highest number of target genes in the Prenatal and Neonatal, Prenatal and Adult, Neonatal, Neonatal and Adolescent, Adolescent and Adult, and Adult groups respectively.
Fig 8.
Validation of novel transcript variants.
(A) Skipping exon 2 in Oatp1b2 (B) Intron retention between exons 1 and 2 in Bsep. Data are visualized by the IGV (Broad Institute). Reads mapped to the gene from a day.2 liver are shown in the examples. (C) A semiquantitative PCR gel electrophoresis of the two Oatp1b2 isoforms and two Bsep isoforms along with Gapdh.
Fig 9.
Examples of the different novel isoform variants.
(A-C) Examples of novel isoforms from the selected genes. The novel isoform track is the first track on top in black. This is followed by the RefSeq Genes track (light blue), the Ensembl Genes track (red) and UCSC Genes track (dark blue). The expression pattern of the novel isoforms over the 12 ages is shown in the embedded graphs. The different features constituting the novel isoforms are highlighted by red circles and labeled as appropriate. Exons are shown as boxes. Arrows indicate the direction of transcription.