FACS sorted Cells were first washed in 0.5% bovine serum albumin in phosphate-buffered saline (BSAPBS, Sigma) solution to avoid potential contamination. After washing, fresh cells were collected in 1.5 ml tubes and resuspended in 45μl lysis buffer (40 mM Tris-HCl, pH 7.5, 10 mM NaCl, 6 mM MgCl2, 6 mM CaCl2, 0.1% NP40) then incubated on ice for 5 min to release the nuclei. DNase I (10,000 U/ml, Roche, 04716728001) was added to the final concentration of 30 U/ml, and incubated at 37 °C for exactly 5 min. DNase I concentrations to achieve optimal smearing sizes may differ for each cell line and therefore should be determined empirically for each cell type. After DNase I digestion, the reaction was terminated immediately by adding 50μl Stop Buffer (40 mM Tris-HCl, pH 7.5, 10 mM NaCl, 0.5% SDS, 10 mM EDTA) containing 1μl Proteinase K (20 mg/ml, Life Technologies). After incubation at 55 °C for 1h, DNA was purified by extraction with phenol–chloroform and precipitated by ethanol in the presence of glycogen (20 mg/ml, Roche, 10901393001) more than 3 hours at −20 °C. The sequence libraries were generated using the KAPA Hyper Prep Kit for the Illumina platform (kk8504), following the manufacturer's instructions. In order to evaluate the library complexity, we chose agarose gel electrophoresis instead of SPRI-beads based methods for size selection. Because DNase I hypersensitive sites are not binary, only the smearing pattern samples were ideal for generating high-quality DNase data. DNA fragments of 150-300bp were extracted from the agarose gel for a second PCR amplification. The DNA was eluted in 15μl of elution buffer (2.5 mM Tris, pH 7.6, 0.05 mM EDTA) and quantified by a Qubit dsDNA HS assay kit (Thermo Fisher Scientific, Q32854) and Agilent High Sensitivity assay kit (Agilent Technologies). The libraries were sequenced on a Hiseq2500 with single-end 50bp reads (Illumina).