Skip to main content
. Author manuscript; available in PMC: 2014 Dec 24.
Published in final edited form as: J Proteomics Bioinform. 2014 Feb 17;7:1000302. doi: 10.4172/jpb.1000302

Figure 4. Relationship between predicted promoter class and ongoing transcription (GRO-seq).

Figure 4

(A and B) Genes categorized based on their predicted promoter class (‘active in both’, ‘exclusively pre-pro-B’, ‘exclusively pro-B ‘, and ‘inactive in both’) are shown in box and whiskers plots representing the quartiles of GRO-seq values plotted versus GRO-seq transcription levels (RPKM on a log10 scale) in pre-pro-B cells (A) and pro-B cells (B). In these representations, all protein-coding genes in RefSeq are considered, including genes with RPKM = 0. (C and D) Graphs showing the relative frequency distribution of GRO-seq transcription levels [log10(RPKM)] for each predicted promoter class (see color key) in pre-pro-B cells (C) and pro-B cells (D). (E and F) Graphs showing the relative frequency distribution of the differential ratio of GRO-seq transcription levels [log10(RPKMpro-B/RPKMpre-pro-B)] for each predicted chromatin class, with the magenta box of (E) magnified in (F). In these representations, frequency was assessed at increments of 0.1 of the log10 converted transcription level (C and D) or calculated differential (E and F), and genes with RPKM = 0 are not included (C–F). All classes demonstrated statistically significant differences (WMW p < 0.0001).