Additional file 1 of Universal NicE-seq for high-resolution accessible chromatin profiling for formaldehyde-fixed and FFPE tissues

Additional file 1: Supp Fig. 1 Optimization of universal NicE-seq (A) A schematic diagram of accessible chromatin labeling using dCTP or 5-mdCTP in the labeling reaction along with biotinylated-dCTP in the nucleotide mix. On-bead and off-bead represented presence of streptavidin magnetic beads for DNA capture and library preparation. (B) FRiP comparison between all 4 methods generated library that map to TSSs (+/-500 bp of TSS) and distal elements (>500 bp from TSS) from HCT116 cells. C and 5mC represents use of dCTP and 5-dCTP in the reaction mix. Supp Fig. 2: Optimization of accessible chromatin sequencing and comparison between UniNicE-seq, ATAC-seq and DNase-seq. (A) IGV screen shot of the normalized read density of the four NicE-seq conditions in HCT116 cells. (B) Distribution of the number of normalized HCT116 NicE-seq reads at transcription start sites (TSS) of human genes and the surrounding 2 Kb (- and +) regions. (C) Pearson correlation of normalized read densities in UniNicE-seq peaks of the 2 technical replicates in HCT116 demonstrating reproducibility. (D) IGV screen shot of the normalized read density of UniNicE-seq (top track), ATAC-seq (middle track) and DNase-seq (bottom track) in HCT116 (F) Overlap of HCT116 peaks called from 15 M unique alignments using UniNicE-seq, ATAC-seq and DNase-seq. Supp Fig. 3: Venn Diagram showing common and cell-type specific UniNicE-seq peaks between the three cell types. (A) HCT116, K562 and MCF7 accessible chromatin regions were analyzed. Peaks are called from 11 million random sampled deduplicated alignment pairs. Supp Fig. 4: UniNicE-seq of mouse T cells cells. (A) IGV screen shot of the normalized read density of the technical duplicates of UniNicE-seq libraries of HCT116 cells at different cell numbers. (B) Pairwise comparison between all Universal NicE-seq reads between different T cell numbers from 500, 5 and 25 K. Pearson’s correlation is indicated. Supp Fig. 5: Comparison between UniNicE-seq, ATAC-seq, Omini ATAC-seq and DNase-seq of mouse kidney cells. (A) Venn diagram of accessible chromatin regions derived from UniNicE-seq, ATAC-seq, Omini ATAC-seq and DNase-seq of mouse kidney cells. (B) Distribution of fold change (FC) values (derived from MACS2) of the common accessible chromatin peaks of UniNicE-seq, ATAC-seq, Omni ATAC-seq and DNase-seq. (C) FRiP score of UniNicE-seq 25 K, 0.5 K and 0.25 K compared with data obtained from ATAC-seq and OmniATAC-seq using 50 K cells, and DNase-seq. (D) Heatmap showing comparison of normalized RPKM of 25 K fixed, 0.5 K nonfixed and 0.25 K mouse kidney UniNicE-seq, 50 K omni ATAC-seq, 50 K ATAC-seq and DNase-seq data at TSS, PolII and random. (E) Similar comparison like (D) along with chromatin features including CTCF, H3K4me3, H3K27Ac and random fragments. Supp Fig. 6: Comparison between accessible chromatin sequences two liver FFPE tissue section (A) Venn diagram demonstrating common accessible regions in two different human 5-10 μm lung normal tissue sections. (B) Pearson’s correlation analysis of total reads between two different human 5-10 μm lung normal tissue sections. (C) Pearson’s correlation analysis of common reads between two different human 5-10 μm lung normal tissue sections demonstrating quality of accessible peaks. (D) Principle Component Analysis and heat map of TSS across fetal and adult tissue for normalized read density of the consensus peaks between the samples from fetal and adult tissue. (E) Heat map of TSS (-/+ 2 kb) between various tissue samples. Supp Fig. 7: UniNicE-seq of normal human liver FFPE tissue section. Pearson’s correlation analysis by pairwise comparison between two different liver samples, R1 and R2. Supp Fig. 8: GC content of HCT116 peaks called from 15 M unique alignment pairs using UniNicE-seq, ATAC-seq and DNase-seq. The last box represents the GC content of random genomic regions sampled from the human reference genome (hg38).附加文件1：补充图1 通用NicE-seq（universal NicE-seq）的优化（A）标记反应中使用dCTP或5-mdCTP，并在核苷酸混合物中加入生物素化dCTP的开放染色质标记示意图。On-bead与off-bead分别代表用于DNA捕获及文库制备的链霉亲和素磁珠的存在状态。（B）4种方法构建的文库在HCT116细胞中比对到转录起始位点（Transcription Start Site, TSS，±500 bp范围）及远端元件（距TSS＞500 bp）的结合分数（FRiP）比较。C与5mC分别代表反应体系中使用dCTP与5-dCTP。 补充图2 开放染色质测序优化及UniNicE-seq、ATAC-seq与DNase-seq的比较（A）HCT116细胞中4种NicE-seq条件的标准化读长密度IGV（Integrative Genomics Viewer）截图。（B）人类基因转录起始位点（TSS）及上下游2 kb区域内标准化后的HCT116 NicE-seq读长数量分布。（C）HCT116细胞中2次技术重复的UniNicE-seq峰内标准化读长密度的皮尔逊相关性分析，验证实验重复性。（D）HCT116细胞中UniNicE-seq（上方轨迹）、ATAC-seq（中间轨迹）与DNase-seq（下方轨迹）的标准化读长密度IGV截图。（F）利用UniNicE-seq、ATAC-seq与DNase-seq对15 M唯一比对读段进行峰识别后得到的峰重叠情况。 补充图3 韦恩图（Venn Diagram）展示3种细胞类型间共有的及细胞类型特异性的UniNicE-seq峰。（A）分析HCT116、K562与MCF7细胞的开放染色质区域，峰识别基于1100万条随机抽样的去重比对读段对。 补充图4 小鼠T细胞的UniNicE-seq实验（A）不同细胞数量的HCT116细胞UniNicE-seq文库技术重复的标准化读长密度IGV截图。（B）500、5与25 K（千）不同细胞数量的通用NicE-seq读段间的两两比较，标注皮尔逊相关系数。 补充图5 小鼠肾细胞的UniNicE-seq、ATAC-seq、Omni ATAC-seq与DNase-seq比较（A）小鼠肾细胞的UniNicE-seq、ATAC-seq、Omni ATAC-seq与DNase-seq得到的开放染色质区域韦恩图。（B）UniNicE-seq、ATAC-seq、Omni ATAC-seq与DNase-seq共有的开放染色质峰的倍数变化（FC，由MACS2分析得到）分布。（C）25 K、0.5 K与0.25 K细胞量的UniNicE-seq的FRiP得分，与使用50 K细胞的ATAC-seq、OmniATAC-seq及DNase-seq数据进行对比。（D）热图展示在TSS、PolII及随机区域中，25 K固定、0.5 K未固定与0.25 K小鼠肾细胞UniNicE-seq、50 K Omni ATAC-seq、50 K ATAC-seq及DNase-seq数据的标准化RPKM（Reads Per Kilobase per Million mapped reads）比较。（E）与（D）类似的比较，同时纳入CTCF（CCCTC结合因子）、H3K4me3（组蛋白H3赖氨酸4三甲基化）、H3K27Ac（组蛋白H3赖氨酸27乙酰化）及随机片段等染色质特征。 补充图6 两份福尔马林固定石蜡包埋（FFPE）肝脏组织切片的开放染色质序列比较（A）两份不同的人类5-10 μm正常肺组织切片中共有的开放染色质区域韦恩图。（B）两份不同的人类5-10 μm正常肺组织切片的总读段间皮尔逊相关性分析。（C）两份不同的人类5-10 μm正常肺组织切片的共有读段间皮尔逊相关性分析，用于验证开放染色质峰的质量。（D）胎儿与成人组织样本的共识峰标准化读长密度的主成分分析（PCA）及TSS热图。（E）不同组织样本的TSS（±2 kb）区域热图。 补充图7 正常人类肝脏福尔马林固定石蜡包埋（FFPE）组织切片的UniNicE-seq实验。两份不同肝脏样本R1与R2间的两两比较的皮尔逊相关性分析。 补充图8 利用UniNicE-seq、ATAC-seq与DNase-seq从15 M唯一比对读段对中识别出的HCT116峰的GC含量分布。最后一个箱线图代表从人类参考基因组（hg38）中抽样得到的随机基因组区域的GC含量。