Genome sequencing and assembly
Total genomic DNA was further fractionated into 10-50 kb fragments with BluePippin to construct libraries following the Nanopore library construction protocol. The libraries were sequenced at the Nextomics Biosciences Company (Wuhan, China) using the GridION X5 sequencer platform (Oxford Nanopore Technologies, UK). Quality-controlled reads were used for de novo assembly using Nextdenovo v. 2.3.051 (https://github.com/Nextomics/NextDenovo). We corrected sequencing errors with NextCorrect, assembled with NextGraph, and polished with NextPolish v. 1.2.424 (Hu et al., 2020). At this stage, the short and long reads were used three times for genome correction. Finally, purge_haplotigs (Roach et al., 2018) was used to remove allelic haplotigs, resulting in the final genome assembly. Benchmarking Universal Single-Copy Orthologs (BUSCO) v. 2.026 (Simão et al., 2015), with 1,350 genes from Embryophyta odb10, was used to evaluate the completeness and accuracy of the assembled genome.