Chip seq libraries

For mRNA-seq 1 μg total RNA was incubated with poly-dT beads to isolate polyadenylated RNA, subjected to fragmentation and cDNA synthesis followed by library preparation (TruSeq RNA Sample Prep kit v2) performed according to the manufacturer's (Illumina, San Diego, CA, USA) instructions. ChIP-seq libraries were constructed according to the manufacturer's instructions (Illumina) as described previously.^{57 (link)} Sequencing was carried out on a HiSeq1500 platform (Illumina).

The DSX tissue was finely ground in liquid nitrogen and chromatin was extracted as described in Li et al.^{69 (link)}. ChIP was performed with ~2 μg anti-H3K4me3 (Merck Millipore #07–473), anti-H3K27me3 (Merck Millipore, ABE44) or anti-RNA Pol II (Merck Millipore, #17–672) antibodies using the protocol described in Adli and Bernstein^{70 (link)}, without ChIP DNA amplification. Illumina ChIP-seq libraries of immuno-enriched and input samples were prepared and sequenced (SE50 reads; 200 bp library insert) by the Vincent L. Coates Genomics Sequencing Laboratory, University of Berkeley, CA. Read quality assessment and mapping to the E. grandis genome v.1.1^{20 (link)} was performed as described in Hussey et al.^{21 (link)}. Strand cross-correlation analysis was performed as described in Landt et al.^{22 (link)}. Significantly enriched regions were identified using the ENCODE Irreproducible Discovery Rate (conservative) method with a threshold of 0.01^{66 (link)} on peaks identified using MACS v.2^{71 (link)}. Peaks overlapping with those identified using a nonspecific antibody^{21 (link)} were discarded. Genes were regarded as modified for each HM if the annotated transcribed regions overlapped a significant ChIP-seq peak to any degree.