The accurate and comprehensive DNase I HS map presented here offers an unprecedented view of open chromatin structure at extremely high resolution. We have shown that both tiled microarrays and high-throughput sequencing are very accurate at identifying DNase I HS sites across the genome and combining these platforms improves the sensitivity, specificity, and the ability to determine the degree of hypersensitivity. We believe this is especially important for correctly calling DNase I HS sites that are more moderately hypersensitive. In future studies, most would find it undesirable and likely cost prohibitive to employ both technologies. While both methods are very accurate at identifying DNase I HS sites, each method has unique benefits and limitations. For example, we have demonstrated that DNase-seq can also be used to detect sub-nucleosome structure, something not possible with current tiling arrays. However, DNase-seq analysis on aneuploid cell lines will be difficult without performing extensive sequencing of the input DNA from each cell type. In contrast, tiled arrays readily normalize for DNA content and thus are suitable for cells with abnormal karyotypes. In addition, while DNase-seq can currently only be used to study the whole genome, tiling arrays can be used for inexpensive validation and for studying smaller targeted regions of the genome. Neither platform is well suited for duplicated sequences such as those found in recent segmental duplications, or for other highly repetitive sequences. Therefore, it is difficult to compare costs of the two technologies, since it depends on the particular application.
DNase I HS maps provide a scaffold on which to combine and analyze data from ChIP-chip/ChIP-Seq and gene expression experiments to better understand complex gene regulation. In our limited study in a single cell type, we are able to show previously undescribed positional relationships between DNase I HS peaks, transcription start sites, and sites of RNA PolII binding. We also describe differences in histone modifications around different categories of HS sites based on their degree of hypersensitivity, their positional relationship to transcription start sites, and the expression level of associated genes.
As similar types of data continue to be produced from different cell types, we anticipate the development of regulatory maps consisting of DNase I HS sites that are characterized by the presence or absence of features such as histone modifications, DNA binding proteins, DNA methylation, nucleosome positions, SNPs, insertions and deletions that collectively explain the transcriptional status of associated genes in particular cell types under particular conditions. In addition, DNase I HS maps can be used to focus computational motif discovery and analyses on those regions of the genome most likely to contain functional binding, a role that evolutionary conservation has not satisfactorily filled (ENCODE, 2007
The data resource presented here should be of particular interest to those studying the biology of CD4+ T cells, the regulation of genes that are expressed in many cell types, and those studying comparative genomics. We have shown that we efficiently identify previously characterized HS sites in these cells, and our data should therefore benefit future research. The further generation of genome-wide DNase I HS maps from a diverse set of normal and diseased human cell types, as well as from those from other species, will continue to reveal how chromatin structure, and underlying primary sequence differences, contribute to cell-type specific gene expression and cell fate decisions.