epiRomics Package Example Dataset

The epiRomics package includes example data delineating human pancreatic islet enhancers between alpha and beta cells.

Human pancreatic islet alpha and beta ATAC- and companion RNA-Seq data were retrieved from GEO accession GSE76268 (Ackermann et al., 2016).

ATAC samples were processed using the ENCODE-DCC ATAC sequencing pipeline, aligning to the hg38 build of the human genome (Harrow et al., 2012; Consortium, 2012; Davis et al., 2018).

Peak calls generated through the pipeline using MACS2 (Zhang et al., 2008) were analyzed downstream through the Bioconductor package DiffBind (Ross-Innes et al., 2012) to identify differentially enriched chromatin regions between the two cell types.

RNA samples were quality controlled using fastp (Chen et al., 2018), and aligned using STAR (Dobin et al., 2013) to hg38. Wiggle files produced by the STAR aligner were merged by cell type using UCSC command line tools.

BigWig files were subsetted using UCSC command line tools (Kent et al., 2010).

ChIP-seq peak calls for human pancreatic islet transcription factors Foxa2, MafB, Nkx2.2, Nkx6.1, and Pdx1 were retrieved from EMBL-EBI E-MTAB-1919 (Pasquali et al., 2014).

Histone modification peak calls for H3K27ac, H3K4me1, H3K27me3, H3K9me3, H3K4me3, H3K36me3 were retrieved from GEO GSE16256 (Bernstein et al., 2010), and H2A.Z from EMBL-EBI E-MTAB-1919 (Pasquali et al., 2014).

All peak calls retrieved as hg19 were lifted over to hg38 using the UCSC liftOver (Kent et al., 2002).

The FANTOM5 human enhancer database (Lizio et al., 2015), ultra-conserved non-coding elements from the UCNE database (Dimitrieva and Bucher, 2012), and the human islet regulome (Miguel-Escalada et al., 2019) were also retrieved and lifted to hg38.