Abstract
The Chinese bahaba (Bahaba taipingensis) is a critically endangered species found in the coastal waters of the South China Sea and East China Sea, which has garnered significant attention due to its ecological, economic, and medicinal value. In this study, we present a high-quality genome assembly of the Chinese bahaba. By employing a combination of PacBio HiFi, ONT ultra-long and Hi-C sequencing data, we assembled 686.86 Mb genome sequences, consisting of 83 contigs with a contig N50 of 30.02 Mb. Then, the initial assembly was further anchored onto 24 chromosomes, of which only 2 contain 3 gaps and 8 have telomeres at both ends. Percentages of 96.5% of complete BUSCOs were achieved, and a total of 22,100 protein-coding genes were annotated. We performed transcriptome analysis across nine tissues (heart, intestine, kidney, liver, muscle, skin, spleen, stomach, swim bladder). The result revealed higher expression levels of col1a1 and col1a2 in swim bladder, which may contribute to massive collagen deposition in this tissue. This high-quality assembly provides a valuable resource for studying the genetic basis of key biological traits (e.g., swim bladder collagen synthesis, otolith development) and will assist future investigations into the conservation genomics of B. taipingensis.
Background & Summary
The Chinese bahaba (Bahaba taipingensis), a member of the Sciaenidae family (Perciformes), is a large-sized fish native to the coastal waters of the South China Sea and East China Sea. It is renowned for its swim bladder (dried as “fish maw”), which is valued in traditional Chinese medicine for treating postpartum hemorrhage, oligoasthenospermia1, and colorectal cancer2. Due to overfishing and habitat degradation3, this species has been listed as Critically Endangered (CR) in the IUCN Red List and a national first-class protected wild animal in China.
Genomics has played a crucial role in elucidating the genetic underpinnings and regulatory mechanisms of key traits in fishes, including growth, development, sex chromosome evolution, and disease resistance. The year 2002 marked a milestone with the completion of the first fish genome assembly of the pufferfish (Fugu rubripes), and since then, over 300 fish genome assemblies have been generated. Notably, genome assemblies for 13 Sciaenidae species (e.g., Larimichthys crocea4, Collichthys lucidus5, Nibea coibor6) have been successfully published, providing foundations for exploring phylogenetic relationships and adaptive traits (e.g., golden body color, sex-determination loci), and molecular breeding of Sciaenidae.
Recent studies on B. taipingensis have focused on artificial reproduction, resource assessment, genetic load and medicinal value7,8. However, genomic resources for this species remain limited: the lack of a high-continuity (near telomere-to-telomere or gap-reduced) genome assembly has hindered in-depth investigations into its key biological traits and conservation genetics.
To address the genomic gap, we integrated Pacific Biosciences (PacBio) HiFi, Oxford Nanopore Technologies (ONT) ultra-long, and Hi-C data to generate a chromosome-level genome assembly of B. taipingensis. We also performed repeat annotation, gene prediction, and transcriptome profiling of 9 tissues. Key results include: (1) a 686.86 Mb assembly with contig N50 of 30.02 Mb, anchored to 24 chromosomes; (2) functional annotation of 22,100 genes and identification of collagen-related genes (col1a1, col1a2) highly expressed in the swim bladder; (3) genome data deposited in public databases for reuse. These resources support research on B. taipingensis conservation, trait evolution, and molecular breeding.
Methods
Sample collection and sequencing
A 1-year-old female B. taipingensis was obtained as fresh dead samples from Guangdong Beluga Whale Marine Biotechnology Co., Ltd. (Guangdong, China) and processed within 24 hours of acquisition to ensure sample quality. Sex was confirmed through paraffin sectioning and hematoxylin-eosin (HE) staining of gonadal tissue. Nine tissues, including heart, intestine, kidney, liver, muscle, skin, spleen, stomach, and swim bladder, were promptly frozen in liquid nitrogen and stored at −80 °C.
For genomic sequencing, DNA extraction, library construction, and sequencing were conducted using muscle tissue. Specifically, for PacBio HiFi sequencing, high-molecular-weight (HMW) DNA was extracted with the Monarch Genomic DNA Purification Kit T3010 (New England Biolabs, USA). A library was then prepared with the SMRTbell Prep Kit 3.0 (PacBio, USA) and sequenced on the PacBio Sequel II platform, yielding 31.59 Gb of HiFi reads. For ONT ultra-long sequencing, HMW DNA was isolated via SDS (sodium dodecyl sulfate)-based lysis and purified with phenol-chloroform. A library was constructed using the Ligation Sequencing Kit SQK-LSK110 (ONT, UK) and sequenced on an ONT PromethION platform, generating 12.62 Gb of ultra-long reads. For Hi-C sequencing, samples were fixed with 1% formaldehyde, digested with MboI, and labeled with biotin-14-dCTP. Paired-end libraries (insert size ~300 bp) were subsequently constructed with the MGIEasy Hi-C Library Prep Set (MGI, China) and sequenced on the BGISEQ-500 platform, producing 60.92 Gb of raw Hi-C data.
For transcriptome sequencing, RNA was extracted from 9 tissues using TRIzol reagent (Invitrogen, USA). RNA libraries were prepared with the MGIEasy RNA Library Prep Set (MGI, China) and sequenced on the BGISEQ-500 platform (paired-end 150 bp reads).
Genome assembly
Contig assembly was performed with Hifiasm (v0.19.5), utilizing PacBio HiFi, ONT ultra-long, and Hi-C data as input. During chromosome construction, the Hi-C technique was applied. Using HiC-Pro9, read pairs were mapped to the genome, with both ends uniquely aligned retained to ensure data quality. Juicer (v1.5)10 and 3D-DNA11 were then employed to construct the chromosome-level genome. Completeness was assessed with BUSCO (v5.3.0) using the actinopterygii_odb10 database12. Base accuracy was verified via read mapping rates (Illumina reads aligned with BWA-MEM v0.7.17).
By combining HiFi, ONT ultra-long and Hi-C data, we assembled 686.86 Mb genome sequences, consisting of 83 contigs with a contig N50 of 30.02 Mb (Table 1). The GC content of the genome is 41.34% (Table 1). The length of contig N50 shows the high quality assemble level of this genome, as it is much larger than most of other Sciaenidae fish. The result of BUSCO shows that we get 3,511 complete BUSCOs, which includes 3,478 complete and single-copy BUSCOs and 33 complete and duplicated BUSCOs. There are also 52 fragmented BUSCOs and 77 missing BUSCOs. There were no obvious false duplications (Table 2). Finally, the initial assembly was further anchored onto 24 chromosomes (Fig. 1a–c; Table S1), with a cc (contig-to-chromosome) ratio of 3.46. Among these chromosomes, 8 exhibited telomeres at both ends, and only 2 chromosomes contained a total of 3 gaps. The consensus quality value (QV) of the assembly was assessed by Merqury (v1.3)13 to be 57.5366 using HiFi reads.
Table 1.
Genome assembly summary.
| statistics | scaffold | contig |
|---|---|---|
| Total Number (#): | 80 | 83 |
| Total length (bp): | 686864734 | 686863234 |
| Gap(N)(bp): | 1500 | 0 |
| Average Length (bp): | 8585809.18 | 8275460.65 |
| N50 Length (bp): | 30022115 | 30022115 |
| N90 Length (bp): | 22828916 | 22828916 |
| Maximum Length (bp): | 35025592 | 35025592 |
| Minimum Length (bp): | 16849 | 16849 |
| GC content: | 41.34% | 41.34% |
Table 2.
BUSCO summary.
| Kind of BUSCO | Number |
|---|---|
| Complete BUSCOs | 3511 |
| Complete and single-copy BUSCOs | 3478 |
| Complete and duplicated BUSCOs | 33 |
| Fragmented BUSCOs | 52 |
| Missing BUSCOs | 77 |
| Total BUSCO groups searched | 3640 |
Fig. 1.
Chromosome construction. (a) Chromosomal contact maps. The blocks represent the contacts between one location and another. The colour reflects the intensity of each contact, with deeper colours representing higher intensity; (b) Chromosome treemap. Each rectangle’s size is proportional to the length of its corresponding contig or chromosome; (c) The circus plot of the genome features. From outer to inner: distribution of DNA TEs, distribution of RNA TEs, gene density, GC content, the 24 chromosomes, each line connects paralogous gene pairs; (d) Functional annotation of the gene set. The set size means the total number of genes annotated by different databases; The vertical bars show the total number of genes present within an annotation or shared between annotations indicated by black dots present below the x-axis of the plot.
Repeat elements prediction
Repeat elements in the genome were identified using multiple tools and databases. Tools such as RepeatModeler (v1.0.9)14, LTR FINDER (v1.05)15, and the TRF tool (v4.04) were used for de novo prediction. Subsequently, RepeatMasker (v3.3.0) was utilized using the RepBase library (v20181026) with the settings -nolow -norna -no_is16, and RepeatProteinMask was employed with parameters -engine ncbi -noLowSimple -pvalue 1e-0417.
A total of 136.51 Mb (19.87%) transposable elements (TEs) were predicted. Among these TEs, there were 10.99% of DNA repeat elements, 8.49% of long interspersed nuclear elements (LINE), 0.95% of short interspersed nuclear elements (SINE), and 3.085% of LTR (Table 3).
Table 3.
Repeat annotation.
| Type | Repbase TEs | Protein TEs | Denovo TEs | Combined TEs | |
|---|---|---|---|---|---|
| Length(bp) | Length(bp) | Length(bp) | Length(bp) | %genome | |
| DNA | 26,238,886 | 530,751 | 62,182,226 | 75,477,616 | 10.989 |
| LINE | 15,496,971 | 11,617,555 | 51,055,201 | 58,297,081 | 8.487 |
| SINE | 2,837,866 | 0 | 4,239,548 | 6,514,011 | 0.948 |
| LTR | 8,312,358 | 3,739,966 | 16,118,161 | 21,189,470 | 3.085 |
| Other | 7,306 | 0 | 0 | 7,306 | 0.001 |
| Unknown | 0 | 0 | 6,493,952 | 6,493,952 | 0.945 |
| Total | 47,801,338 | 15,886,597 | 125,354,525 | 136,506,912 | 19.874 |
Gene structure prediction and functional annotation
Genome structure analysis incorporated homology-based, transcriptome-based, and de novo prediction methods. For homology-based annotation, protein sequences from 8 teleost species: Cynoglossus semilaevis, Danio rerio, Gadus morhua, Gasterosteus aculeatus, L. crocea, Oreochromis niloticus, Oryzias latipes, and Takifugu rubripes, were aligned to the genome with GENEWISE (v2.4.0)18. Transcript annotation involved assembling transcripts from RNA-Seq data with TRINITY (v2.1.1)19, aligning them to the genome using PASA20, and predicting ORFs with TransDecoder. De novo gene prediction was carried out using AUGUSTUS (v2.5.5)21, GlimmerHMM (v3.0.4)22, and GENSCAN (v2.1)23, followed by integration using Glean. A total of 22,100 protein-coding genes were identified in the final annotated gene set.
Functional annotation of gene sets was conducted using 4 databases, including NR, KEGG24, SwissProt25, and TrEMBL, with homologous proteins identified via BLASTp using an E-value cut-off of 1E-526. Domain-based annotation was conducted using InterProScan (v4.7)27.
Transcriptome data Analysis
Transcriptome data were processed using Fastp (v0.23.2) for quality control, Hisat2 (v2.1.0) for accurate transcript alignment, and featureCounts (v2.0.6) for efficient quantification of transcript abundance. These raw counts were subsequently converted into Transcripts Per Million (TPM) values for normalization and comparative analysis. A weighted gene coexpression network analysis (WGCNA) was performed using the R (v4.3.3) package WGCNA (v1.72.5), following published procedure28.
In this study, we investigated the collagen distribution in swim bladder and the other 8 tissues. There were 42 genes related to collagen that had a TPM value greater than 10 in at least one sample, as shown in Fig. 2a. The genes col1a1, col1a2, col6a6 and col10a1 were highly expressed in swim bladder. The WGCNA was performed using 18,116 genes from 18 samples and 38 modules were identified (Fig. 2b). It was found that the four collagens were all included in the blue module. Then the GO and KEGG analysis of blue module was conducted, the results showed that the top enriched GO terms included ion binding, extracellular matrix, growth and collagen trimer (Fig. 2c).
Fig. 2.
Gene expression profiles. (a) Heatmap of collagen genes. (b) Relationship between modules and samples, where the colour bar indicates the correlation value from low (green) to high (red). (c) GO enrichment of genes in the blue module.
Ethics statement
The animal study was reviewed and approved by the Ethics Committee of the Yellow Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences.
Data Records
The sequencing data and the genome assembly have been deposited in National Center for Biotechnology Information (NCBI) under the Bioproject accession number PRJNA1344354. The sequencing data including PacBio HiFi, ONT ultra-long, and Hi-C reads, are available in the NCBI Sequence Read Archive (SRA) under the respective accession numbers SRR3577861729, SRR3577861630, and SRR3581010031. The genome assembly has been deposited in NCBI GenBank under accession number GCA_054948935.1 (YSFRI_Bta_1.1)32. The genome annotation files were submitted to FigShare33.
Technical Validation
The quality of the genome assembly was assessed in terms of continuity, completeness, and accuracy. The assembly has a total length of 686.86 Mb with a contig N50 of 30.02 Mb and a scaffold N50 of 30.02 Mb. The chromosome assembly shows strong interaction patterns in Hi-C contact maps, and 8 chromosomes are gap-free with telomeres identified at both ends. BUSCO assessment revealed 3,511 complete BUSCOs (96.5% completeness), with 3,478 single-copy and 33 duplicated genes. The GC content is 41.34%. Mapping rates of Illumina reads to the assembly exceeded 99%, indicating high accuracy. Repeat annotation identified 136.51 Mb (19.87%) of repetitive sequences, consistent with other Sciaenidae genomes (e.g., L. crocea: 26.13%34; N. coibor: 18.31%6; Table S2).
Supplementary information
Author contributions
S.C. conceived the project and revised the manuscript. M.L. analyzed data and drafted the manuscript. Z.G. and W.X. supervised the data and revised the manuscript. Z.G. collected and prepared the sequencing samples. All authors reviewed the manuscript.
Funding
This work was supported by Key Research and Development Project of Shandong Province (2023ZLYS02), Agricultural Biobreeding Major Project (2023ZD0405501), Taishan Scholar Climbing Project of Shandong Province, Central Public-interest Scientific Institution Basal Research Fund, CAFS (2023TD20), the earmarked fund for China Agriculture Research System (CARS-47-G03).
Data availability
The sequencing data and the genome assembly have been deposited in National Centre for Biotechnology Information (NCBI) under the Bioproject accession number PRJNA1344354. The genome annotation files were submitted to FigShare (https://doi.org/10.6084/m9.figshare.30547922).
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Ming Li, Wenteng Xu, Zehao Ma, Zhihong Gong.
Supplementary information
The online version contains supplementary material available at https://doi.org/10.1038/s41597-026-07501-y.
References
- 1.Zhao, X. et al. Peptides from the croceine croaker (Larimichthys crocea) swim bladder attenuate busulfan-induced oligoasthenospermia in mice. Pharm Biol60(1), 319–325 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Suo, H. et al. Induction of apoptosis in HCT-116 colon cancer cells by polysaccharide of Larimichthys crocea swim bladder. Oncol Lett9(2), 972–978 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Agency, X. N. Chinese scientists launch project to protect critically endangered fish, in China Daily. (2025).
- 4.Mu, Y. et al. An improved genome assembly for Larimichthys crocea reveals hepcidin gene expansion with diversified regulation and function. Commun Biol1, 195 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cai, M. et al. Chromosome assembly of Collichthys lucidus, a fish of Sciaenidae with a multiple sex chromosome system. Sci Data6(1), 132 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Yekefenhazi, D. et al. Chromosome-level genome assembly of Nibea coibor using PacBio HiFi reads and Hi-C technologies. Sci Data9(1), 670 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yan, L. et al. Early Growth and Developmental Characteristics of Chinese Bahaba (Bahaba taipingensis). Fishes9(8), 329 (2024). [Google Scholar]
- 8.Cui, R. et al. Phased genome assemblies reveal haplotype-specific genetic load in the critically endangered Chinese Bahaba (Teleostei, Sciaenidae). Mol Ecol33(4), e17250 (2024). [DOI] [PubMed] [Google Scholar]
- 9.Servant, N. et al. HiC-Pro: an optimized and flexible pipeline for Hi-C data processing. Genome Biol16, 259 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Durand, N. C. et al. Juicer Provides a One-Click System for Analyzing Loop-Resolution Hi-C Experiments. Cell Syst3(1), 95–98 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dudchenko, O. et al. De novo assembly of the Aedes aegypti genome using Hi-C yields chromosome-length scaffolds. Science356(6333), 92–95 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Simao, F. A. et al. BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics31(19), 3210–3212 (2015). [DOI] [PubMed] [Google Scholar]
- 13.Rhie, A. et al. Merqury: reference-free quality, completeness, and phasing assessment for genome assemblies. Genome Biol21(1), 245 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Flynn, J. M. et al. RepeatModeler2 for automated genomic discovery of transposable element families. Proc Natl Acad Sci USA117(17), 9451–9457 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Xu, Z. & Wang, H. LTR_FINDER: an efficient tool for the prediction of full-length LTR retrotransposons. Nucleic Acids Res35(Web Server issue), W265–268 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tarailo-Graovac, M. & Chen, N. Using RepeatMasker to identify repetitive elements in genomic sequences. Curr Protoc Bioinformatics, Chapter 4: p. Unit 4 10 (2009). [DOI] [PubMed]
- 17.Bao, W., Kojima, K. K. & Kohany, O. Repbase Update, a database of repetitive elements in eukaryotic genomes. Mob DNA6, 11 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Doerks, T. et al. Systematic identification of novel protein domain families associated with nuclear functions. Genome Res12(1), 47–56 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Grabherr, M. G. et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nature Biotechnology29(7), 644–652 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Haas, B. J. et al. Improving the Arabidopsis genome annotation using maximal transcript alignment assemblies. Nucleic Acids Res31(19), 5654–5666 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Stanke, M. et al. AUGUSTUS: ab initio prediction of alternative transcripts. Nucleic Acids Res34(Web Server issue), W435–439 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Majoros, W. H., Pertea, M. & Salzberg, S. L. TigrScan and GlimmerHMM: two open source ab initio eukaryotic gene-finders. Bioinformatics20(16), 2878–2879 (2004). [DOI] [PubMed] [Google Scholar]
- 23.Burge, C. & Karlin, S. Prediction of complete gene structures in human genomic DNA. J Mol Biol268(1), 78–94 (1997). [DOI] [PubMed] [Google Scholar]
- 24.Kanehisa, M. & Goto, S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res28(1), 27–30 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bairoch, A. & Apweiler, R. The SWISS-PROT protein sequence data bank and its supplement TrEMBL. Nucleic Acids Res25(1), 31–36 (1997). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bairoch, A. et al. The Universal Protein Resource (UniProt). Nucleic Acids Res33(Database issue), D154–159 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mitchell, A. L. et al. InterPro in 2019: improving coverage, classification and access to protein sequence annotations. Nucleic Acids Res47(D1), D351–D360 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics9, 559 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.NCBI Sequence Read Archive.https://identifiers.org/ncbi/insdc.sra:SRR35778617 (2026).
- 30.NCBI Sequence Read Archive.https://identifiers.org/ncbi/insdc.sra:SRR35778616 (2026).
- 31.NCBI Sequence Read Archive.https://identifiers.org/ncbi/insdc.sra:SRR35810100 (2026).
- 32.NCBI GenBank.https://identifiers.org/ncbi/insdc.gca:GCA_054948935.1 (2026).
- 33.Li, M. Chromosome-Level Genome Assembly and Annotation of the Chinese Bahaba (Bahaba taipingensis). figshare 10.6084/m9.figshare.30547922 (2025). [DOI] [PMC free article] [PubMed]
- 34.Chen, B. H. et al. The sequencing and de novo assembly of the Larimichthys crocea genome using PacBio and Hi-C technologies. Sci Data6, 188 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Citations
- Li, M. Chromosome-Level Genome Assembly and Annotation of the Chinese Bahaba (Bahaba taipingensis). figshare 10.6084/m9.figshare.30547922 (2025). [DOI] [PMC free article] [PubMed]
Supplementary Materials
Data Availability Statement
The sequencing data and the genome assembly have been deposited in National Centre for Biotechnology Information (NCBI) under the Bioproject accession number PRJNA1344354. The genome annotation files were submitted to FigShare (https://doi.org/10.6084/m9.figshare.30547922).