Draft genome of Roscoea Debilis, the first genome in the alpine ginger Roscoea (Zingiberaceae)

Abstract

Objectives

Roscoea is a Sino-Himalayan alpine genus in pantropical family Zingiberaeae. As traditional Tibetan medicinal plants, many species of this genus are threatened by digging, logging, land clearance, grazing and climate change. Roscoea debilis is an endemic species in the Hengduan Mountains with a narrow distribution range. In this study, the assembled and annotated genome of Roscoea was presented in order to furnish significant resources for comparative and functional genomic investigations. The first complete reference genome of Roscoea is expected to shed light on research on conservation and evolutionary biology.

Data description

A chromosome-level genome of 1601.04 Mb was obtained for R. debilis by combining Illumina short reads (107.28 Gb) and PacBio Hi-Fi reads (64.08 Gb), achieving high-quality sequencing coverage of roughly 67 × and 40 ×. The assembly was additionally assisted by 271.65 Gb Hi-C data (169 ×), which resulted in a contig N50 of 136.17 Mb and a scaffold N50 of 90.48 Mb. Benchmarking Universal Single-Copy Orthologs (BUSCO) assessment results revealed that most of the core embryophyta genes (98.7%) in the BUSCO dataset (embryophyta_odb10) were successfully identified. Additionally, 96.44% of the genomic sequences were accurately mapped onto twelve pseudochromosomes.

Objective

Roscoea is an alpine genus of the Sino-Himalayan distribution, characterized by a unique pattern of disjunctive distribution from the Hengduan Mountains to the Himalayas, and is known for its orchid-like flowers and diverse mating systems [1,2,3]. Roscoea plants have been traditional Tibetan medicines since ancient times in their place of origin [4,5,6], and also exhibit the potential to be used for horticultural and ornamental purposes [7]. In addition, Roscoea plants have a variety of mating systems, including specialized outcrossing, facultative outcrossing, and early selfing [8,9,10]. They are an ideal group for exploring the evolution of mating systems and their response to environmental changes in mountainous regions. Coupled with their high-altitude distribution characteristics, they deserve more attention in both practical application and ecological significance. However, many species of this genus are threatened by digging, logging, land clearance, grazing and climate change.

A high-quality reference genome is useful for various aspects of plant ecology and biology research; however, there is currently no complete genome material available for Roscoea. Roscoea debilis is an endemic species in the Hengduan Mountains, with a narrow distribution range and a unique mating system of early selfing [11]. In this study, the high-quality genome of R. debilis was assembled and annotated using a combination of Illumina sequencing and PacBio Hi-Fi sequencing. To the authors’ knowledge, this is the first complete reference genome of Roscoea, which can provide important basic data for future research on Roscoea.

Data description

Leaf samples of R. debilis were obtained from an individual located in Tonghai, Yunnan, Southwest China (102.69272°N, 24.073225°E). The genome DNA was extracted from the leaves, and the total RNA was extracted from all tissues, to form four sequencing libraries are formed. Subsequently, Hi-Fi long-read whole genome sequencing (WGS) was performed through PacBio Sequel II sequencers, and short-read WGS, RNA-seq, and Hi-C sequencing were carried out with Illumina NovaSeq 6000 sequencers. Under the MGI platform, a 150 bp paired-end mode was applied for both short WGS and RNAseq. The cross-linked and lysed cells for Hi-C library were digested using Dnp II restriction enzyme. Hi-Fi sequencing generated approximately 64.08 Gb of long-read WGS data (data file 1). Illumina produced about 107.28 Gb of short-read WGS data (data file 2). RNA-seq produced around 6.15 Gb of data (data file 3) for auxiliary commentary. Hi-C sequencing generated approximately 271.65 Gb of data (data file 4).

Following the process of sequencing, the genome survey of R. debilis was conducted using BBmap [12] with Illumina short-read data. The findings suggested that the R. debilis genome was diploid and highly heterozygous, with a heterozygosity rate of 0.7%, and an estimated genome size of around 1.65Gb. The genome was subsequently assembled using Hi-Fiasm [13] with clean Hi-Fi long reads. Hi-Fi reads were assembled using Purge_dups [14] to eliminate redundant sequences. The assembled sequence was further polished with Illumina short reads and NextPolish [15], yielding a preliminary assembled genome version (contig level). In order to achieve a genome at the chromosomal level, Hi-C reads were aligned to major contigs using chromap [16]. Pair reads that were valid were adopted for the subsequent assembly. YaHS [17] was employed for Hi-C scaffolding. The resulting genome contigs achieved a combined length of 1601.04 Mb, which closely matched the estimated size. The contig N50 was 90.48 Mb, and 96.44% of the contig sequences were aligned to 12 pseudochromosomes. The total scaffolds length was 1544.01 Mb, and the scaffold N50 was 136.17 Mb. The Benchmark Universal Single-Copy Orthologous (BUSCO) [18] evaluation, based on the embryophyte_odb10 dataset, demonstrated that 98.7% of genes were successfully identified, indicating a high level of integrity in the assembly.

RepeatModeler [19], RepeatMasker [20], and RepBase + Dfam databases were utilized for homologous and de novo prediction of repeat sequences. A total of 1,307,034,346 bp of repeats were predicted, accounting for 84.65% of the whole genome, with 878,082,544 bp being LTR.

For gene structure prediction, Hisat2 [21] was adopted for RNA-seq based prediction, miniprot [22] was used for homology prediction, and augustus [23] was employed for de novo prediction. Eventually, EVM [24] was integrated to obtain the final annotation results. A total of 30,745 genes were predicted.

Limitations

Despite the high integrity of the draft genome of R. debilis through contig N50, scaffold N50, and BUSCO results, there are still 28 gaps in the assembly of this version. Therefore, elaborate research on Roscoea requires more complete genome assembly. In the future, the R. debilis genome can be assembled to T2T (Telomere-to-Telomere) level using either ONT (Oxford Nanopore Technologies) sequencing or deeper Hi-Fi sequencing, which will yield a greater amount of genomic information.

Table 1 Overview of all data files/data sets