Skip to main content
. 2014 Oct 31;15(10):494. doi: 10.1186/s13059-014-0494-z

Figure 2.

Figure 2

CIITA BIs in relation to gene location, CIITA module and gene function. (A) Distribution of CIITA BIs showing relative distances from the TSS of genes. (B) Enrichment of genes that were differentially expressed in treated monocytes within two groupings of CIITA target genes, one associated with CE marks (CIITA BIs overlapping RFX5) and the other CIITA BIs without RFX5. The empirical cumulative distribution functions (eCDFs) for each grouping is shown with a 95% confidence interval. A group of CIITA targets containing a larger than expected number of differentially expressed genes will produce an eCDF that lies above the line corresponding to absence of enrichment. (C) Enrichment for all CIITA BIs (overlapping and non-overlapping with RFX5) associated with known and novel CIITA target genes in treated monocytes. (D) Box plot showing averaged rate of MATCH-PWM binding site scores for a CIITA module of RFX5, CREB and NF-Y motifs (co-factors of the classical CIITA enhanceosome) in promoters of known CIITA target genes and those identified in this study (denoted as novel targets where CIITA BIs localised within 10 kb 5′ of the TSS). For known and novel target genes, 150 bp intervals centred on CIITA peak summits were analysed. A control set comprised of 1,000 randomly selected segments is included for comparison. Distributions of the three sets were compared using Mann-Whitney-Wilcoxon test, showing that these are significantly different, between the control set and known CIITA targets (P = 2.2 × 10-16) or novel targets (P = 1.5 × 10-11), and between known and novel targets (P = 2.7 × 10-7).