Library preparation was performed on 0.5-1x105 cells by Tagment DNA enzyme and buffer Small kit (Illumina), as previously described,88 (link) with minor modifications, and sequenced by 2x150 Illumina HiSeq.
Paired-end FASTQ reads were trimmed with trimmomatic (v 0.36) using the Nextera adapter library. Trimmed reads were then mapped on the hg19 version of the human genome using bwa-mem with default settings and duplicates were marked using samblaster (v 0.1.21). Duplicates, blacklisted regions (ENCODE) and reads aligning with mitochondrial DNA were removed using samtools (v 0.1.19). Broad peaks were called using MACS2 (v. 2.1.1; --shift −100 --extsize 200 --broad). The list of called peaks from each sample was used to assess the differentially open chromatin regions through the DiffBind package from the R statistical environment using edgeR and DESeq2 methods (FDR<0.01 and FC > 2 for in vitro data, FC > 4 for ex vivo data, where fold change values were larger). Matching results obtained by the two methods were used to define the list of differentially accessible regions that were in common across both DC-10 and DC. In case of groups with less than 3 samples (AHRinhDC-10 experiment) or more than 3 samples (ex vivo DC-10 experiment), only EdgeR or DESeq2 results were considered, respectively.
Paired-end FASTQ reads were trimmed with trimmomatic (v 0.36) using the Nextera adapter library. Trimmed reads were then mapped on the hg19 version of the human genome using bwa-mem with default settings and duplicates were marked using samblaster (v 0.1.21). Duplicates, blacklisted regions (ENCODE) and reads aligning with mitochondrial DNA were removed using samtools (v 0.1.19). Broad peaks were called using MACS2 (v. 2.1.1; --shift −100 --extsize 200 --broad). The list of called peaks from each sample was used to assess the differentially open chromatin regions through the DiffBind package from the R statistical environment using edgeR and DESeq2 methods (FDR<0.01 and FC > 2 for in vitro data, FC > 4 for ex vivo data, where fold change values were larger). Matching results obtained by the two methods were used to define the list of differentially accessible regions that were in common across both DC-10 and DC. In case of groups with less than 3 samples (AHRinhDC-10 experiment) or more than 3 samples (ex vivo DC-10 experiment), only EdgeR or DESeq2 results were considered, respectively.