Figure 1.

An Expression Atlas of Human Organ Development

(A) Description of the dataset. The dots mark the developmental stages sampled in each organ (median of two replicates), and its colors reflect the colors used throughout the figures to represent each organ.

(B) Modules in the gene co-expression network (number of genes in each module in parentheses). The left panel shows the correlations between each module and the expression levels in each organ or developmental time (full developmental profiles in Figure S1A). The color intensity and the size of the circles are proportional to the correlation coefficients. A positive correlation with developmental time (first column) means higher expression late in development, and a negative correlation means higher expression early in development. The middle panel shows the fraction of genes in each module that correspond to TFs, RBPs, developmentally dynamic lncRNAs, and poorly studied protein-coding genes (i.e., genes with three or fewer publications). The right panel shows examples of overrepresented diseases (FDR <1%, hypergeometric test). Table S1 lists the top five biological and disease enrichments (FDR <1%) for each of the 32 modules. The modules are ordered vertically by decreasing number of genes. Module 0 (bottom) includes genes not assigned to any of the other modules.