Fig. 3

Gene expression profile and sample-to-sample distance relationships. a Heatmap representing the expression levels (read counts) of the differentially expressed transcripts (FDR < 0.01, FC > 2) across samples: 0, 3, 6, and 12 h (“3 h 1”, “3 h 2”, “6 h 1”, “6 h2”, “12 h 1” and “12 h 2”). Read counts were mean-centered and scaled, and the resulting values were represented in a green red false color scale (green = underexpressed; red = overexpressed). Columns and rows were clustered by the Spearman correlation coefficient among samples and transcripts, respectively. b Principal component analysis (PCA) showing how the samples are grouped based on transcripts CPM (counts per million) values at each sample. Top plot shows the first and second components (53.22% of variance), and bottom plot shows the second and third components (31.81% of variance)