4.3. ChIP-Seq Data Processing and Super Enhancer Analysis

The mapping of ChIP-seq data (Table 1) to hg19 was performed using bowtie2 [50]. H3K27ac peak calling was performed using MACS1.4 with default parameters [51]. Peak calling was separately performed for each sample. Peaks that could not be identified in each colorectal cancer sample and peaks that appeared within the region surrounding ±2.5 kb of transcriptional start sites were excluded from any further analysis. Afterwards, the H3K27ac peaks of the 7 individual samples (7 published colorectal cancer H3K27ac CHIP-seq datasets from ENCODE [16] and a published study [17]) were merged into a single set of peaks. Super enhancers were identified using the rank ordering of super enhancers (ROSE) algorithm [14]. For colorectal-cancer-specific enhancers, we removed all peak regions that contained any overlap with a peak detected in each normal colon region.

ChIP-seq data sets.