Pacbio sequel

Sequence coverage was calculated using the previously established genome size estimate for T. brassicae of 246 Mbp (Johnston et al. 2004 (link)). Library preparation and sequencing was performed by Novogene Bioinformatics Technology Co., Ltd., (Beijing, China). For Illumina sequencing, gDNA was used to construct one paired-end (PE) library according to the standard protocol for Illumina with an average insert size of 150 bp and was sequenced using an Illumina HiSeq 2000 (Illumina, San Diego, USA). For Single Molecule Real Time (SMRT) sequencing, gDNA was selected for optimal size using a Blue Pippin size selection system (Sage Science, Beverley, USA) following a standard library preparation. The library was then sequenced on a PacBio Sequel (Pacific Biosciences, Menlo Park, USA) with 16 SMRT cells.

Genomic DNA was extracted using the QIAGEN DNaesy Plant Mini Kit according to the manufacturer’s protocols. The extracted DNA molecules were sequenced by PacBio Sequel (Pacific Biosciences of California, Menlo Park, CA, USA) platforms. The CLR reads were assembled using mecat2 (20,190,226) with default parameters. Draft genome was corrected by arrow and pilon. Polished contigs were anchored to chromosome by Hi-C reads. First, Hi-C reads were mapped to the polished H. rotundifolia genome using BWA (bwa-0.7.17) and Lachesis with default parameters [63 (link)]. Paired reads with mate mapped to a different contig were used to do the Hi-C-associated scaffolding. Lachesis was further applied to cluster, order, and orient the clustered contigs. Two methods were combined to identify the repeat contents in H. rotundifolia genome, homology-based, and de novo prediction. The repeat contents found by these two methods were merged by repeat masker. Protein-coding genes of the H. rotundifolia genome were predicted by three methods, including ab initio gene prediction, homology-based gene prediction, and RNA-Seq-aided gene prediction (Supplementary Methods 2 for genome sequencing, assembly, and annotation details).

Genomic DNA was extracted using the modified CTAB method and sequenced using PacBio Sequel (Pacific Biosciences) combined with Illumina HiSeq 2500 sequencing for correction. The genome size and heterozygosity of safflower were determined using flow cytometry and k‐mer frequency analysis. The 17 k‐mer distribution displayed one major peak. The k‐mer depth of 33 was determined as the main peak depth of the k‐mer frequency distribution using 150‐bp Illumina paired‐end reads (35 Gb).
Briefly, we performed de novo assembly using Canu (version 1.3; Koren et al.,2017 (link)) and Falcon (Chin et al.,2016 (link)), and found Canu (N50 = 16.43 Mb with 368 contigs) is better than Falcon (1.40 Mb with 3195 contigs). Then, the draft genome from Canu assembly was furthered assembled into scaffold whit Hi‐C. Then, we used the GPM pipeline (Zhang et al.,2016 (link)) to fill gaps of 12 superscaffolds (N50 = 14.17, 213 contigs; Superscaffold N50 = 88.21Mb) from Hi‐C with Falcon contigs after filtering then unanchored short contigs. Finally, the consensus sequence was corrected using the arrow method implemented in the SMRTLink and further polished with long reads using Pilon (version 1.22; Walker et al.,2014 (link)) to get a final version (N50 = 21.23 Mb with 128 contigs).

Genomic DNA was fragmented and sheared using a g-TUBE (Covaris Inc., Woburn, MA, USA) and then purified using AMPure PB magnetic beads (Beckman Coulter Inc. Bread, CA, USA). A 10 μL library was prepared using the PacBio DNA template Prep Kit 1.0 (for 3–10 kb). SMRTbell templates were annealed using PacBio DNA/Polymerase Binding Kit P6. Libraries were sequenced on PacBio SMRT Cell using the PacBio Sequel (Pacific Biosciences) sequencing platform (Macrogen; Seoul, Korea), yielding 490,426 reads (average read length, 9,180; N50 value, 16,358; total read length, 4,502,006,607 bp). Some reads that were fully contained in other reads provide no extra information with respect to constructing the genome; these were filtered out. After quality control, 81,788 reads were recovered (average length, 5,709; N50 value, 9,479; total read length, 466,969,904 bp). The filtered reads were used to assemble the genome using HGAP4 v4.0, with default options. The genome sequence was circularized using the Circlator algorithm and annotated using the RAST server or the NCBI Prokaryotic Genome Annotation Pipeline (PGAP). The Circos program was used to draw the chromosome map. Secondary metabolites were predicted using antiSMASH. CRISPR loci were detected using CRISPRFinder. The genome sequence was deposited in the NCBI BioProject under accession number PRJNA557809.

The samples were loaded on the PacBio Sequel (Pacific Biosciences) following the protocol titled “Protocol for loading the Sequel.” The recipe for loading the instrument was generated by the Pacbio SMRTlink software v 5.0.0. Libraries were prepared using Sequel chemistry kits v 2.1, SMRTbell template kit 1.0 SPv3, magbead v2 kit for magbead loading, sequencing primer v3, and SMRTbell cleanup columns v2. Libraries were loaded at between 4 and 8pM.