Extended Data Figure 1: Validation of STAR ChIP–seq in mES cells.

a, A brief schematic of STAR ChIP–seq. Based on a previously described ChIP method²⁸, micrococcal nuclease (MNase) is used to fragment native chromatin. Nucleosomes with certain histone modifications are immunoprecipitated using specific antibodies. Proteinase K is added for protein digestion and is heat-inactivated. Shrimp alkaline phosphatase (rSAP) is directly added to the reaction for 3′ DNA end repair, which is required for the poly-C addition in the subsequent TELP library preparation. After heat-inactivation of rSAP, the resulting DNA is subjected to the TELP library preparation as previously described³ without purification. Briefly, poly-C tailing is conducted on denatured single-strand DNA using dCTP and terminal deoxynucleotidyl transferase (TDT). Biotin-labelled anchor primer (indicated by ‘B’) containing poly-G is used for second strand DNA extension. The products are captured by magnetic streptavidin beads. After an adaptor ligation to the opposite end of poly-C, the double-stranded DNA is amplified by primers (P1 and P2) containing Illumina adaptor sequences. The resulting library is ready to be sequenced. b, A snapshot of the UCSC browser view shows enrichment of STAR ChIP–seq for histone modifications H3K4me3, H3K27ac and H3K27me3 using various numbers of mES cells. Data sets from mouse ENCODE are also shown for comparison³⁵. c, The Pearson correlation coefficients showing the comparison between STAR ChIP–seq of H3K4me3 using various numbers of mES cells and conventional ChIP–seq of H3K4me3 from either this study using 1 × 10⁶ mES cells or data from ENCODE³⁵. d, A bar chart showing the percentages of conventional mESC H3K4me3 total/promoter peaks recaptured by STAR ChIP using various numbers of mES cells. e, Box plots showing the H3K4me3 enrichment determined by conventional ChIP–seq at peaks that are either recaptured or missed by STAR ChIP–seq.