Background & Summary

The global decline in freshwater biodiversity has become a pressing concern. A recent assessment by the International Union for Conservation of Nature (IUCN) reported that 3,086 out of 14,898 assessed freshwater fish species are currently threatened with extinction (IUCN, 2023). This loss of biodiversity is driven by multiple factors, including global climate change1,2, human-induced impacts such as river exploitation and hydroelectric development, and the spread of invasive species3,4. Invasive taxa have been associated with negative ecological consequences, such as predation on native small fish, habitat destruction, and substantial reductions in both biodiversity and biomass5,6,7,8,9. In China, findings from the third national assessment of the Red List of freshwater fishes indicate that 22.3% of species are facing extinction risks10. In response, various conservation strategies have been introduced, including the designation of freshwater fish conservation zones, implementation of fishing moratoriums, and execution of restocking programs11,12. Nonetheless, the overall efficacy of these measures remains to be systematically assessed.

Decorus tungting, also known as Bangana tungting (Nichols, 1925), classified as a vulnerable species by both the Red List and FishBase, exemplifies the conservation challenges associated with declining freshwater biodiversity. This large-bodied indigenous cyprinid, endemic to China, historically supported an annual catch exceeding 160 tons in the 1940s and 1950s13. However, field investigations conducted between 2006 and 2010 revealed a drastic population collapse throughout its native range, mainly due to overfishing and river system modifications13. Following this sharp decline, the species was classified as endangered.

The species is now largely restricted to the Yuan River National Aquatic Germplasm Resources Reserve (YUFRR), established in 2010 with strict no-fishing policies. Substantial progress has been made in artificial breeding and restocking efforts, with 8.65 million juveniles bred and released into native rivers between 2007 and 2021 through designated release sites14. A recent survey confirmed the effectiveness of protective measures in maintaining juvenile populations but also noted the rarity of natural recruitment, indicating that the wild population remains non-self-sustaining15. Given the precarious status of D. tungting, there is an urgent need to characterize its genetic diversity, adaptive capacity, and germplasm characteristics at the molecular level, so as to support conservation strategies such as genetically informed restocking, assessment of inbreeding and genetic load, and conservation genetic management16. A high-quality reference genome would provide an important foundation for conservation-oriented molecular analyses, as it would substantially facilitate downstream genetic assessments essential for implementing these strategies17.

Genomics approaches have become increasingly integrated into conservation biology, providing powerful tools to assess demographic history, population structure, and adaptive potential. Advances in high-throughput sequencing technologies have facilitated the genome sequencing of various fish species. However, within the Cyprinidae family, genomic efforts have largely focused on model organisms or species of economic significance, leaving ecologically significant but genetically understudied species like D. tungting largely neglected.

The taxonomy of the genus Bangana sensu lato was recently revised by Zheng18 and Song et al.19 based on molecular phylogenetic and morphological evidence, leading to the establishment of the genus Decorus. This genus currently includes D. decora, D. lemassoni, D. rendahli, D. tungting, D. xanthogenys, and an undescribed species (D. sp.). Despite their ecological and taxonomic importance, no genome assemblies have been published for any Decorus species, leaving a critical gap in the molecular resources needed for evolutionary and conservation studies. The lack of genomic resources has constrained studies in D. tungting, making it challenging to evaluate genetic diversity, identifying adaptive signals and deleterious variants, or delineating evolutionarily significant units (ESUs), which limits the development of effective conservation strategies.

A high-quality reference genome can enable a suite of genomic applications under the framework of genomics-assisted conservation, including population genomic analysis, adaptive variation mapping, and marker-assisted conservation breeding. Such approaches have shown promise in improving conservation outcomes in endangered or commercial fishes such as Oncorhynchus masou formosanu20, Anguilla japonica21, and Nannoperca australis22, and are increasingly adopted in conservation planning for non-model species23.

In the present study, we generated a high-quality chromosome-level genome assembly of Decorus tungting using a combination of Illumina NovaSeq. 6000, PacBio long-read sequencing, and high-throughput chromosome conformation capture (Hi-C). Gene annotation was performed through a comprehensive approach integrating de novo prediction, homology-based methods, and RNA-Seq evidence. We also conducted comparative genomic analyses with closely related Cyprinidae species to investigate the evolutionary history of D. tungting. This work represents the first chromosome-level genome assembly for the genus Decorus, providing a fundamental genomic resource for studies on genetic diversity, adaptive evolution, and conservation of this vulnerable species.

Methods

Ethics statement

The methods involving animals in this study adhered to the Laboratory Animal Management Principles of China. All experimental protocols were approved by the Ethics Committee of Hunan Fisheries Science Institute.

Sample collection

Living specimens of Decrous tungting were collected from Tianling Company Limited, Huaihua, China (Fig. 1a). Fish were placed under anesthesia using tricaine methanesulfonate (MS222), after which white muscle tissue was dissected, immediately immersed in liquid nitrogen, and stored at −80 °C until DNA extraction.

Fig. 1

Overview of the Decorus tungting genome assembly. (a) Representative morphological features of D.tungting, including its streamlined fusiform body, subterminal mouth with thick lips, and prominent sensory pores on the snout50. (b) 17-mer frequency distribution of Illumina short reads estimated using GenomeScope. The main peak corresponds to the homozygous peak, used to estimate genome size (~1.12 Gb), heterozygosity, and repeat content.

DNA Extraction and genome sequencing

Genomic DNA was extracted from muscle tissue using the proteinase-K–chloroform method. DNA quality and integrity were assessed by 1% agarose gel electrophoresis, and fragment size distribution was evaluated using a Bioanalyzer or PFGE to confirm the presence of high-molecular-weight DNA for long-read sequencing. High-quality DNA was used to construct libraries following the 20 kb PacBio template protocol with BluePippin size selection. DNA was sheared into ~20 kb fragments using a g-TUBE, then end-repaired and ligated to SMRTbell adapters. Libraries were purified, quality-checked, and mixed with primers and polymerase before sequencing on the PacBio Sequel II platform. Sequencing was conducted in CCS mode, generating high-fidelity HiFi reads for accurate genome assembly.

RNA Extraction and transcriptome sequencing

Total RNA was extracted from white muscle tissue using TRIzol. RNA integrity and contamination were assessed by 1% agarose gel electrophoresis, and concentration was measured using a Qubit fluorometer. Poly(A) + mRNAs were fragmented using fragmentation buffer and reverse-transcribed into cDNA with random hexamer primers. The cDNA was end-repaired, A-tailed, and ligated to sequencing adapters to prepare libraries with an average insert size of 350 bp. Libraries were sequenced in paired-end mode on the Illumina NovaSeq. 6000 platform, generating transcriptome data for genome annotation.

De Novo Genome assembly and quality assessment

De novo genome assembly of D.tungting was performed using HiFi reads generated on the PacBio Sequel II platform. Genomic DNA was prepared with the SMRTbell Express Template Prep Kit 2.0 according to standard protocols. Sequencing was performed using Circular Consensus Sequencing (CCS v6.0.0) with a minimum of three passes, yielding 20.4 Gb of HiFi reads with an N50 of 13,539 bp, representing approximately 102 × sequencing coverage of the estimated genome size.

Prior to assembly, genome size, heterozygosity, and repeat content were estimated via 17-mer analysis using 9.5 Gb of quality-trimmed Illumina reads (TrimGalore v0.6.124, Q < 20). Jellyfish v2.3.025 and GenomeScope v1.026 were used to analyze k-mer distributions. HiFi reads were then assembled using hifiasm v0.16.127 with diploid optimization (-D 50 -N 500), yielding a final genome assembly size of 1.14 Gb, which is in good agreement with the k-mer-based estimate of ~1.12 Gb (Fig. 1b). Two rounds of polishing were performed with Pilon v1.2328, using BWA v0.7.17-aligned Illumina29 reads to correct indels and base errors. Assembly completeness was evaluated using BUSCO v5.2.230 with the Actinopterygii_odb10 dataset.

The final assembly resulted in a total genome size of 1.14 Gb across 117 scaffolds (Table 1). At the contig level, the assembly exhibited high continuity, with an N50 of 6.04 Mb, an N70 of 3.71 Mb, and an N90 of 1.41 Mb, while the longest contig reached 23.48 Mb. Following scaffolding, the scaffold-level assembly was further improved, achieving an N50 of 44.81 Mb, an N70 of 41.47 Mb, and an N90 of 37.52 Mb, with the longest scaffold measuring 74.21 Mb. The GC content of the assembled genome was 36.43%, and a total of 620 gaps corresponding to 0.31 Mb were detected. The 17-mer frequency analysis validated the high accuracy of the assembly (Fig. 1b).

Hi-C Library preparation and chromosome scaffolding

Muscle tissues were collected for Hi-C library construction. Genomic DNA was extracted and crosslinked with formaldehyde to stabilize chromatin interactions. Crosslinked DNA was digested using DpnII, end-labeled with biotinylated nucleotides, ligated to form chimeric junctions, then purified and fragmented to ~350 bp. Hi-C libraries were prepared using the GrandOmics Hi-C kit and sequenced on the Illumina NovaSeq. 6000 platform in paired-end mode.

Raw Hi-C reads were processed using Juicer v1.631. Contigs from the primary assembly were clustered, ordered, and oriented into chromosome-scale scaffolds using 3D-DNA v18092232. Manual curation was conducted with Juicebox v1.8.833 to refine scaffold structure. After Hi-C scaffolding, a total of 25 chromosomes were reconstructed, anchoring 96.42% of the genome assembly (Table 2), which is consistent with a previous cytogenetic study reporting the diploid karyotype of D. tungting as 2n = 5034.The assembled chromosomes ranged in length from ~35.5 Mb to 74.2 Mb, containing between 753 and 1,702 predicted genes. Chromosomal GC content ranged from 35.76% to 37.74%. The high-resolution Hi-C contact map confirmed successful chromosomal scaffolding, with well-defined diagonal interaction patterns indicative of chromosome-scale continuity and integrity (Fig. 2a). These genome-wide features, including gene density, repeat content, and GC composition, were visualized using a Circos plot (Fig. 2b), providing an overview of structural organization.

Fig. 2

The diagonal interaction signals indicate scaffold continuity and chromosomal integrity. (a) Hi-C contact heatmap showing chromosome-level assembly of D. tungting. (b) Circos plot of distribution of the genomic elements in Decorus tungting. For the outer to inner regions, each circle represents the gene density (A), repeative element density (B), GC content (C).

Annotation of repeat elements

Repeat elements (REs) in the genome were annotated using the Extensive de-novo TE Annotator (EDTA, v1.9.9) pipeline35,36,which integrates multiple de novo transposable element (TE) identification tools. Known repeats from the Repbase database, including cyprinid-specific entries, were also incorporated. The identified repeats were masked using RepeatMasker v4.1.137, and annotations were cross-validated with Repbase.

Repetitive elements comprised approximately 50.28% of the D. tungting genome, corresponding to ~574.36 Mb (Fig. 3). DNA transposons were the predominant repeat class. Among them, the hAT superfamily contributed the largest fraction, representing 14.95% of the genome. The CACTA and Mutator families accounted for 4.14% and 2.33%, respectively. Helitrons were also prevalent, occupying 9.94% of the genome. Tandem repeats, identified independently using TRF, accounted for 16.11% of the genome. LINEs and SINEs constituted 3.52% and 0.41%, respectively. Within LINEs, the L2, L1, and Rex subfamilies represented 1.81%, 1.08%, and 0.51% of the genome, respectively. SINEs were primarily derived from 5S rRNA and tRNA sequences, contributing 0.19% and 0.13%. A full summary of repeat composition, including LTRs, TIRs, Helitrons, LINEs, SINEs, and low-complexity regions, is provided in Table 3, offering a comprehensive view of the transposable element landscape in the D. tungting genome.

Fig. 3

Abundance and classification of repetitive elements in the D. tungting genome.

Gene annotation and functional assignment

Quality control of raw RNA-Seq reads was performed using FastQC v0.11.938, followed by adapter trimming and filtering with Trimmomatic v0.3939 (Q < 20, length ≥ 50 bp). Cleaned reads were used for transcript assembly to support gene model prediction. Gene prediction was carried out using the MAKER pipeline v3.01.0440, which integrates transcriptomic evidence, homologous protein sequences, and ab initio gene models. HISAT2 v2.2.041 was used to align RNA-Seq reads to the genome to generate transcript alignments. Protein sequences from related species, e.g., Ctenopharyngodon idella, Danio rerio, Gasterosteus aculeatus, Labeo rohita, and Megalobrama amblycephala, were retrieved from public databases and included as homologous evidence. In the first round of MAKER, transcript and protein evidence were used with parameters est2genome = 1 and protein2genome = 1. Gene models with Annotation Edit Distance (AED) < 0.1 were selected to train ab initio predictors (AUGUSTUS42 and SNAP43), which were used in the second round of MAKER to generate additional gene predictions. Genes potentially associated with hAT transposons were further predicted using GENSCAN44. Functional annotation of predicted protein-coding genes was conducted using BLASTP (e-value < 1e−5) against SwissProt, TrEMBL, and KEGG. InterProScan v5.5245 was used to identify conserved domains and assign Gene Ontology (GO) terms.

A total of 24,835 protein-coding genes were predicted in the D. tungting genome (Table 4), comprising 251,015 exons and spanning approximately 73.59 Mb in total. The average number of exons per gene was 10.11, and 2,129 genes were identified as single-exon genes. Functional annotation was successfully assigned to all predicted genes based on at least one public database. Specifically, 93.10% of genes were linked to KEGG pathways, 88.50% matched InterPro entries, 87.31% had GO terms, and 87.80% showed similarity to UniProt records. Additionally, 60.48% of genes were classified into KOG categories. A Venn diagram summarizing annotation overlap across NR, KEGG, InterPro, UniProt, and KOG is shown in Fig. 4. Most genes were annotated in three or more databases, reflecting broad functional characterization. Complete annotation statistics are summarized in Tables 4, 5, detailed results for each annotation database are available in the public repository Figshare[46](https://www.nature.com/articles/s41597-025-06017-1#ref-CR46 “Xie, X. et al. Gene Annotation and Functional Assignment of Decorus tungting. https://doi.org/10.6084/m9.figshare.29875199

.“).

Fig. 4

Venn diagram for functional annotations of Decorus tungting protein-coding genes with the public database, including NR, KEGG, Interpro, Uniprot, and KOG.

Data Records

The genome sequencing and assembly data for Decorus tungting have been deposited in the China National Center for Bioinformation (CNCB) under the BioProject accession PRJCA040001[47](https://www.nature.com/articles/s41597-025-06017-1#ref-CR47 “China National Center for Bioinformation. https://ngdc.cncb.ac.cn/bioproject/browse/PRJCA040001

.“). The PacBio HiFi sequencing reads are available under the BioSample accession SAMC5089076[48](https://www.nature.com/articles/s41597-025-06017-1#ref-CR48 “China National Center for Bioinformation. https://ngdc.cncb.ac.cn/biosample/browse/SAMC5089076

.“). The assembled genome has been deposited in the Genome Warehouse (GWH) under accession number GWHGDIL00000000.1[49](https://www.nature.com/articles/s41597-025-06017-1#ref-CR49 “China National Center for Bioinformation. https://ngdc.cncb.ac.cn/gwh/Assembly/95114/show

.“). Functional annotation results, including KEGG, GO, Pfam, UniProt/Swiss-Prot, and KOG classifications, are available from the Figshare repository[46](https://www.nature.com/articles/s41597-025-06017-1#ref-CR46 “Xie, X. et al. Gene Annotation and Functional Assignment of Decorus tungting. https://doi.org/10.6084/m9.figshare.29875199

.“). All datasets are publicly accessible through the NGDC portal and Figshare. These resources serve as foundational references for comparative genomics, evolutionary analysis, and molecular breeding studies of cyprinid fish.

Technical Validation

To ensure assembly quality, we evaluated the raw sequencing data from PacBio HiFi, Illumina PE150, and Hi-C PE150 before assembly. Illumina and Hi-C datasets had Q30 values above 91%, and sequencing depth reached ≥ 16 × for PacBio and Illumina and ~96 × for Hi-C, indicating high base quality and adequate coverage for reliable contig construction and chromosome scaffolding (Table 6). Assembly completeness was assessed using BUSCO v5.2.2 with the actinopterygii_odb10 dataset (n = 3,640 conserved single-copy orthologs). In genome mode, 97.1% of BUSCOs were complete, comprising 95.0% single-copy and 2.1% duplicated genes, while 1.2% were fragmented and 1.7% were missing. In protein mode, 92.7% of BUSCOs were complete (89.4% single-copy, 3.3% duplicated), and 6.7% were missing (Table 7). These results support the completeness of the genome and the accuracy of the annotated gene set.

Chromosome-scale scaffolding was achieved using Hi-C data, which anchored 99.31% of the genome to 25 chromosomes. The scaffold N50 reached 44.81 Mb, with the longest chromosome measuring 74.21 Mb. The L50 and L70 values were 11 and 17, respectively, indicating high contiguity and efficient chromosomal anchoring (Table 8). Additionally, repeat analysis revealed that 48.31% of the genome is composed of repetitive sequences, consistent with values reported in other cyprinid genomes. These combined metrics indicate a high-quality assembly suitable for downstream comparative and functional genomic analyses.

Data availability

The genome sequencing and assembly data for Decorus tungting have been deposited in the China National Center for Bioinformation (CNCB) under BioProject accession PRJCA04000148. The PacBio HiFi sequencing reads are available under BioSample accession SAMC508907649. The assembled genome is deposited in the Genome Warehouse (GWH) under accession GWHGDIL00000000.1. Functional annotation datasets are available at Figshare[46](https://www.nature.com/articles/s41597-025-06017-1#ref-CR46 “Xie, X. et al. Gene Annotation and Functional Assignment of Decorus tungting. https://doi.org/10.6084/m9.figshare.29875199

.“).

Code availability

All software and pipelines used for data processing were executed according to the manuals and protocols of the bioinformatics software cited above, and the parameters are clearly described in the Methods section. If no detailed parameters are mentioned for a software, the default parameters were used. The version of the software has been described in Methods. No custom code was used in this study for the curation or validation of the dataset.

References

Aziz, M. S. et al. Decline in fish species diversity due to climatic and anthropogenic factors in Hakaluki Haor, an ecologically critical wetland in northeast Bangladesh. Heliyon 7 (2021). 1.

Barbarossa, V. et al. Threats of global warming to the world’s freshwater fishes. Nat Commun 12, 1701 (2021).

Article ADS PubMed PubMed Central Google Scholar 1.

Barbarossa, V. et al. Impacts of current and future large dams on the geographic range connectivity of freshwater fish worldwide. Proceedings of the National Academy of Sciences 117, 3648–3655 (2020).

Article ADS Google Scholar 1.

Su, G. et al. Human impacts on global freshwater fish biodiversity. Science (1979) 371, 835–838 (2021).

Google Scholar 1.

Yalçın Özdilek, Ş., Partal, N. & Jones, R. I. An invasive species, Carassius gibelio, alters the native fish community through trophic niche competition. Aquat Sci 81, 29 (2019).

Article Google Scholar 1.

Lusk, S., Lusková, V. & Hanel, L. Alien fish species in the Czech Republic and their impact on the native fish fauna. Folia Zool Brno 59, 57–72 (2010).

Article Google Scholar 1.

Strayer, D. L. Alien species in fresh waters: ecological effects, interactions with other stressors, and prospects for the future. Freshw Biol 55, 152–174 (2010).

Article Google Scholar 1.

Castaldelli, G. et al. Introduction of exotic fish species and decline of native species in the lower Po basin, north‐eastern Italy. Aquat Conserv 23, 405–417 (2013).

Article Google Scholar 1.

Wang, J., Chen, L., Tang, W., Heino, J. & Jiang, X. Effects of dam construction and fish invasion on the species, functional and phylogenetic diversity of fish assemblages in the Yellow River Basin. J Environ Manage 293, 112863 (2021).

Article PubMed Google Scholar 1.

Cao, L., Shao, W., Yi, W. & Zhang, E. A review of conservation status of freshwater fish diversity in China. J Fish Biol 104, 345–364 (2024).

Article PubMed Google Scholar 1.

Zheng, Y., Zhang, H., Niu, Z. & Gong, P. Protection efficacy of national wetland reserves in China. Chinese Science Bulletin 57, 1116–1134 (2012).

Article ADS Google Scholar 1.

Liang, Y. & Zhuang, S. Biodiversity in China: challenges, efforts and prospects. China Economic J 17, 26–39 (2024).

Article Google Scholar 1.

Bian, W., Li, C. & Liang, Z. Biological characteristic and resource dynamic of Sinilabeo decorus tungting. Journal of Hydroecology 32, 67–73 (2011).

Google Scholar 1.

Li, S. et al. Histological Study of the Development of the Digestive System of Sinilabeo decorus Tungting Larvae. Journal of Fishery Sciences of China 1033–1043 (2022). 1.

Tian, L. et al. Investigation on the Distribution of Bangana Tungting in Yuanshui Unique Fish Species National Aquatic Germplasm Resources Reserve Using Environmental DNA Technology. Ecol Evol 14, 1–9 (2024).

Article Google Scholar 1.

Arthington, A. H., Dulvy, N. K., Gladstone, W. & Winfield, I. J. Fish conservation in freshwater and marine realms: status, threats and management. Aquat Conserv 26, 838–857 (2016).

Article Google Scholar 1.

Ali, I. Biotechnology in Environmental Conservation: Genetic Tools for Biodiversity Preservation. Frontiers in Biotechnology and Genetics 1, 147–165 (2024).

Google Scholar 1.

Zheng, L. P., Chen, X. Y. & Yang, J. X. Molecular phylogeny and systematic revision of Bangana sensu lato (Teleostei, Cyprinidae). Journal of Zoological Systematics and Evolutionary Research 57, 884–891 (2019).

Article Google Scholar 1.

Song, W. et al. E-cadherin maintains the undifferentiated state of mouse spermatogonial progenitor cells via β-catenin. Cell Biosci 12, 141 (2022).

Article PubMed PubMed Central Google Scholar 1.

Christensen, K. A. et al. Masu salmon species complex relationships and sex chromosomes revealed from analyses of the masu salmon (Oncorhynchus masou masou) genome assembly. G3: Genes, Genomes, Genetics 15, jkae278 (2025).

Article PubMed Google Scholar 1.

Wang, H. et al. A chromosome-level assembly of the Japanese eel genome, insights into gene duplication and chromosomal reorganization. Gigascience 11, giac120 (2022).

Article PubMed PubMed Central Google Scholar 1.

Pavlova, A. et al. Planning and implementing genetic rescue of an endangered freshwater fish population in a regulated river, where low flow reduces breeding opportunities and may trigger inbreeding depression. Evol Appl 17, e13679 (2024).

Article PubMed PubMed Central Google Scholar 1.

Bernos, T. A., Jeffries, K. M. & Mandrak, N. E. Linking genomics and fish conservation decision making: a review. Rev Fish Biol Fish 30, 587–604 (2020).

Article Google Scholar 1.

Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. journal 17, 10–12 (2011).

Article Google Scholar 1.

Marçais, G. & Kingsford, C. A fast, lock-free approach for efficient parallel counting of occurrences of k-mers. Bioinformatics 27, 764–770 (2011).

Article PubMed PubMed Central Google Scholar 1.

Ranallo-Benavidez, T. R., Jaron, K. S. & Schatz, M. C. GenomeScope 2.0 and Smudgeplot for reference-free profiling of polyploid genomes. Nat Commun 11, 1432 (2020).

Article ADS PubMed PubMed Central Google Scholar 1.

Cheng, H., Concepcion, G. T., Feng, X., Zhang, H. & Li, H. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nat Methods 18, 170–175 (2021).

Article PubMed PubMed Central Google Scholar 1.

Walker, B. J. et al. Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS One 9, e112963 (2014).

Article ADS PubMed PubMed Central Google Scholar 1.

Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. bioinformatics 25, 1754–1760 (2009).

Article PubMed [PubMed Central](http://www.ncbi.nlm.nih.gov/p