Figure 2
Characterization of the genomic environment of KRAB/KAP1-docking integrants. LV and MLV TrapSil integrants were split in repressing (REP) and non-repressing (NREP) groups according to the effect of KRAB/KAP1 recruitment on the trapped promoters. The genomic environment of the different proviral integrant groups was analyzed for the indicated genomic features by ROC curve analysis. This method serves to calculate the relative abundance of a given genomic feature around the integrants of a group for specific intervals. The resulting values are depicted in color-coded heat maps. 1 indicates that the specific feature is enriched in integrants, 0 means that it is depleted. Relative abundance scores of repressing and non-repressing integrants were compared for each trapping vector subtype and the statistical method used included the non-central chi-square test (** p < 0.01; *** p < 0.001). The different genomic feature categories tested were: "gene density", with all of the Refseq annotated genes; the "highly expressed genes" and "expressed genes" group, including genes expressed in the top 1/16th, or the top 1/2 of all genes measured in a transcriptional profiling analysis; " start/end", including the distance to the nearest transcriptional start (TSS) or stop site; "gene start", including the distance to the nearest TSS; "gene size" was the average size of the targeted genes and was only analyzed for intragenic integrants; "GC content", included the density of GC nucleotides, which are more abundant in gene-rich regions; "CpG density", contained the frequency of CpG dinucleotides, mostly present at promoters; "DNAse HS sites", included the number of DNAse I hypersensitive sites, frequently associated with gene regulatory regions.