Unbiased genome-scale identification of cis-regulatory modules in the human genome by GRAMc

Abstract

Although significant advances have been made toward functionally identifying human regulatory elements, existing genome-scale reporter methods preferentially detect either enhancers or promoters. Here we develop GRAMc, a highly reproducible unbiased Genome-scale Reporter Assay Method for cis-regulatory modules (CRMs). GRAMc combines the versatility of traditional reporter constructs and the scalability of DNA barcode reporters, and unites the complementary advantages of several currently available high-throughput reporter assays. We demonstrate that GRAMc can reliably measure cis-regulatory activity of nearly 90% of the human genome in 200 million HepG2 cells with randomly fragmented ~800bp inserts. By using the GRAMc-identified CRMs, we show that CRMs identified in one cell type are useful for predicting gene regulatory programs not only within that cell type but also between cell types or conditions separated in time and space. In addition, the GRAMc-identified CRMs support the hypothesis that SINE/Alu elements are rich sources of regulatory evolution. Finally, the observation that the majority of experimentally identified regulatory elements do not overlap with computationally predicted elements underscores the necessity of an efficient and unbiased genome-scale reporter assay.