Abstract
Homatula variegata is a small benthic loach from the upper Yangtze and adjacent basins with aquaculture and ornamental value but no reference genome. We present a near telomere-to-telomere (T2T) chromosome-level assembly built from PacBio HiFi, Oxford Nanopore ultra-long, Illumina short reads, and Hi-C. The 641.26-Mb genome resolves 24 chromosomes as single contigs (contig N50, 24.40 Mb). Hi-C confirms chromosome-length scaffolding; we detect 24 putative centromeres and 20 terminal telomeric tracts, with 22 chromosomes gap-free and two containing one gap. Annotation identifies 24,479 protein-coding genes, 93% functionally assigned, and 27.13% repetitive content dominated by DNA transposons. Quality assessments show high completeness (BUSCO, 96.48% complete) and base-level accuracy consistent with k-mer and read-mapping metrics. To our knowledge this is the first near T2T-level reference for any loach (Cobitoidei), filling a key gap in Cypriniformes genomics. This resource will enable comparative and population genomics, illuminate adaptation to montane stream habitats, and support selective breeding, conservation, and aquaculture of this native species.
Subject terms: Genome, Agricultural genetics
Background & Summary
Homatula variegata (syn. Paracobitis variegatus) is a small benthic fish belonging to the order Cypriniformes and subfamily Nemacheilinae. It is distributed across multiple freshwater systems in China, including the main and tributary streams of the upper Yangtze River in Sichuan and Chongqing, southern Shaanxi, the Bailong River in Gansu, and the Jinsha and Nanpan Jiang rivers in Yunnan1,2. H. variegata inhabits mid-elevation montane streams with pebble substrates and moderate flow. It reaches a maximum total length of approximately 13.9 cm and feeds primarily on benthic invertebrates, detritus, and small fishes, consistent with the ecological traits of Nemacheilinae loaches3. Owing to its palatable flesh, high nutritional content, and vibrant coloration, the species holds economic and ornamental value, particularly as a native ornamental fish with aquaculture potential4.
Research to date has focused on its biology3 and genetic structure5,6. However, no chromosome-scale nuclear genome has been reported for H. variegata. The absence of a high-quality reference genome limits investigations into its evolutionary divergence, local adaptation, population genomics, and the genetic basis of ecologically and economically relevant traits. T2T genome assemblies – gap-free chromosome sequences spanning from one telomere to the other – have recently become attainable thanks to innovations in long-read sequencing and assembly algorithms7. This complete coverage is crucial for capturing genomic regions that were previously unresolved in draft genomes, such as highly repetitive telomeric and centromeric DNA and segmental duplications, thereby revealing novel genes and structural variations that underpin evolution and important8. High-quality T2T references have now been reported for a few fish species (e.g., common carp/koi and Asian seabass) – demonstrating the feasibility and value of gap-free assemblies in non-model organisms9,10. However, no such resource existed for loaches (suborder Cobitoidei), leaving a significant knowledge gap in this lineage. Addressing this gap is a primary goal of our study, as a T2T reference genome for H. variegata would provide a foundation to explore its genomic diversity, adaptive evolution in montane stream environments, and to inform selective breeding, conservation, and aquaculture efforts for this ornamental species.
In this study, we present the first near T2T nuclear genome assembly of H. variegata, generated using a hybrid approach combining PacBio HiFi long reads, Oxford Nanopore ultra-long reads, and Hi-C chromatin conformation capture. A total of 24 chromosomes were assembled, with a genome size of 641.26 Mb and an N50 value of 24.40 Mb (Fig. 1). This assembly fills a critical gap in genomic resources for Nemacheilinae and represents the first near T2T genome for any species within the suborder of Cobitoidei. The genome provides a valuable reference for advancing genetic improvement, resource management, and aquaculture development of H. variegata as a native ornamental and economic species.
Fig. 1.
Circos plot summarizing major genomic features of the assembled genome. From the outside inward: the 24 chromosomes (Chr1-Chr24) arranged clockwise; (I) repeat coverage per 100-kb window (purple, value = fraction of window overlapped by any TE); (II) LTR coverage per 100-kb window (blue); (III) gene density (grey-blue, number of genes per Mb) calculated from gene models; and (IV) GC content (red) computed as (G + C)/(A + T + G + C) within each 100-kb window.
Methods
Ethics statement
All procedures involving H. variegata complied with institutional and national regulations. The protocol was approved by Science and Technology Ethics (Review) Committee of Leshan Vocational and Technical College (approval ID: 202403).
Sample collection
A live adult H. variegata (≥15 g) male was captured by licensed fishers from the Ya’an reach of the Qingyi River (upper Yangtze, China) on 21 August 2024. The fish was rinsed with pre-chilled PBS (4 °C), surface-dried with sterile gauze, and kept on ice. Venous blood (≥0.5 mL) was drawn from the caudal vein using a sterile 1 mL syringe, transferred into pre-chilled EDTA tubes, gently inverted to mix, and immediately snap-frozen in liquid nitrogen. The same blood sample was used for Oxford Nanopore Technologies (ONT) ultra-long read, PacBio HiFi, Illumina whole-genome shotgun (Illumina WGS) and high-throughput chromosome conformation capture (Hi-C) library preparations.
Nucleic-acid extraction and QC
Genomic DNA from whole blood (for Illumina/ONT/PacBio/Hi-C)
High-molecular-weight genomic DNA (gDNA) was extracted from EDTA-anticoagulated whole blood using the Kurabo QuickGene DNA whole blood kit S (DB-S, 40321300101) following the manufacturer’s instructions. Concentration was measured by Qubit Fluorometer (Invitrogen, Qubit 3.0/4.0; dsDNA HS Assay), purity by NanoDrop 2000/One (Thermo), and integrity by 0.8% agarose gel electrophoresis (Tanon EPS600; HE-120 tank). Library preparation proceeded only when total mass ≥6 µg (per PacBio single-library requirement), concentration ≥ 50 ng/µL, OD260/280 = 1.7–2.2, OD260/230 = 1.7–2.5, and HMW smear ≥ 23 kb with minimal degradation.
Total RNA from whole blood (for RNA-seq)
Total RNA was isolated from EDTA blood using TRIzol, followed by column cleanup (Qiagen RNeasy) with on-column DNase I. Quantity (NanoDrop; Qubit RNA HS), integrity (PerkinElmer LabChip GX Touch HT or Agilent Bioanalyzer) and purity were assessed. Libraries were constructed only if RIN ≥ 7 and 28S/18S ≥ 1.8.
Library construction and sequencing
Illumina short-insert WGS
A paired-end library with approximately 350 bp insert size was constructed from genomic DNA using the VAHTS™ Fg DNA Library Prep Kit for Illumina (ND617-02, Vazyme/Yeasen), following enzymatic fragmentation, end repair, A-tailing, adaptor ligation, bead-based size selection, and limited-cycle PCR. The library was quantified with Qubit, and the fragment size distribution evaluated by Qsep-400 (peak 430–530 bp, mean 420–580 bp, single-peak profile). Sequencing was performed on an Illumina NovaSeq X Plus platform using NovaSeq™ X Series 25B reagent kits (PE150). Adapter sequences were: 5′-AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC-3′ and 5′-AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT-3′.
Raw reads were filtered using fastp v0.23.4 (with default settings)11, and read quality summaries generated by FastQC v0.12.1 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). A total of 39.41 Gb of raw paired-end data were generated, corresponding to an estimated sequencing depth of approximately 61.5 × for the ~641.3 Mb genome. Base-level quality metrics were excellent, with Q20 ≥ 99.92% and Q30 ≥ 99.30%, and GC content measured at 39.96%. Post-filtering, clean reads were used to estimate genome properties: genome size ≈ 641.3 Mb, heterozygosity ≈ 0.20%, and repeat content ≈ 36.53% (Fig. 2a).
Fig. 2.
Genome profiling and chromosome-scale scaffolding of H. variegata. (a) 21-mer frequency distribution of Illumina short reads with the GenomeScope2 fit, estimating a genome size of ~641.3 Mb, heterozygosity of ~0.20%, and repeat content of ~36.53%. (b) KR-normalized Hi-C contact heatmap (DpnII-based; Juicer/3D-DNA) of the final assembly, revealing 24 chromosome-length linkage groups with strong on-diagonal signal and minimal inter-chromosomal noise, consistent with correct scaffolding.
ONT ultra-long read sequencing
High-molecular-weight (HMW) gDNA was used to construct ONT ultra-long libraries using the ONT Ultra-long DNA Sequencing Kit V14 (SQK-ULK114), following the manufacturer’s protocol: transposase tagmentation (room temperature 5 min; 75 °C 5 min), adaptor ligation (room temperature 45 min), DNA precipitation and resuspension in 80 µL Elution Buffer with ≥ 24 h of dissolution. Libraries were sequenced on a PromethION 48 using R10.4.1 flow cells (FLO-PRO114M), with runs controlled by MinKNOW software for 72 h.
Basecalling was performed using the MinKNOW-integrated basecaller, or Dorado v0.5.3 (SUP mode) when offline. Adapter trimming utilized Porechop v0.2.4 (--discard_middle) (https://github.com/rrwick/Porechop), followed by filtering with Filtlong v0.2.1 (https://github.com/rrwick/Filtlong), and QC summaries via NanoPlot v1.41.612.
A total of 26.45 Gb of raw ONT reads were generated. After adapter trimming, removal of short and low-quality reads, 22.49 Gb of clean data remained, with metrics including N50 read length of 100 kb, N90 of 39 kb, mean read length of 78 kb, maximum read length of 1,120 kb, and an average per-read quality score of Q12. These ultra-long reads were essential for spanning complex repeat regions and achieving T2T assembly completeness, in keeping with established long-read assembly practice.
PacBio HiFi (CCS)
SMRTbell libraries were prepared using SMRTbell Prep Kit 3.0 (PacBio, 102-182-700), purified with AMPure PB and SMRTbell cleanup beads, and sequenced on PacBio Revio using REVIO SPRQ Polymerase Kit (103-496-900), REVIO SPRQ Sequencing Plate (103-504-900) and REVIO SMRT Cell Tray (102-202-200). CCS generation used Pacbio SMRT Link v13.0 (ccs; --min-passes 3 --min-rq 0.99).
A targeted data output of 42.0 Gb was planned for the project. Sequencing of the single sample yielded ~40.82 Gb of high-quality CCS data, encompassing approximately 2.68 million reads. The reads had a mean length of 15.25 kb, an N50 of 15.34 kb, and a maximum read length of 61.19 kb. The CCS N50 surpassed the QC threshold of ≥10 kb, confirming its suitability for de novo assembly. Moreover, sampling of ~10% CCS reads against the NCBI NT database showed 18.46 mapped reads (13.01% to Cyprinus carpio, 2.20% to Danio rerio). We emphasize that these best-hit species labels are influenced by database representation and conserved regions, and are not interpreted as evidence of contamination or species identity. To confirm sample identity, we extracted the assembly contig that contains the mitochondrial COI (cox1) region and searched it against NCBI nt. The top hit was Homatula laxiclathra (query coverage = 100%, identity = 98.97%), supporting assignment to genus Homatula (Nemacheilidae).
Hi-C libraries
Hi-C libraries were constructed from the blood sample using a DpnII-based protocol: formaldehyde crosslinking, cell lysis, DpnII digestion, end-repair with biotin-dNTPs, proximity ligation, crosslink reversal, purification, fragmentation to ~300–700 bp, and streptavidin bead capture. Libraries were quantified by Qubit and sized by Agilent 2100/Qsep-400, then sequenced on NovaSeq X Plus (PE150).
Adapters and low-quality bases were trimmed with fastp v0.23.411. Reads were aligned to the draft assembly with BWA-MEM2 v2.2.1 (−5 -SP)13, valid pairs identified with HiC-Pro v3.1.014, and.hic matrices produced with Juicer v1.7.615 for visualization.
The Hi-C dataset comprised 69.79 Gb of clean data, with Q30 ≥ 91.8%, meeting project quality benchmarks. Library complexity was high, evidenced by > 47.2% of reads spanning valid restriction enzyme junctions, consistent with typical high-quality Hi-C libraries. These data provided robust chromosomal contact information, essential for accurate scaffolding and correction of assembly structure (Fig. 2b).
Genome profiling
Genome size, heterozygosity and repeat content were estimated from Illumina clean reads using Jellyfish v2.3.016 and GenomeScope217. K-mer analysis (k = 21) based on a 350-bp Illumina library estimated a haploid genome size of ~641.3 Mb, repeat content of ~36.53%, and heterozygosity of ~0.20%.
De novo assembly and polishing
Primary contigs were assembled with hifiasm v0.19.818 using HiFi reads as the primary input and ONT ultra-long reads as an auxiliary input in HiFi + UL hybrid mode (--ul) with otherwise default parameters. Long-read mappings used minimap2 v2.2619. Haplotypic duplications were removed using purge_dups v1.2.620 to generate a collapsed assembly for downstream polishing and scaffolding. Two rounds of short-read polishing were performed with NextPolish v1.4.1 (Illumina)21. To monitor exogenous contamination during assembly, we randomly sampled ~10% of the PacbBio CCS/HiFi reads (n = 267,000) and queried both the read subset and assembly contigs against the NCBI nt database using BLASTN v2.13.022 (-evalue 1e-25 -max_target_seqs. 5). Taxonomic matches were inspected at higher ranks to identify obvious non-target signals and any contigs showing strong matches to clearly unrelated taxa, which would trigger re-evaluation prior to scaffolding. The final assembly spans 641.26 Mb and is anchored to 24 chromosomes (Fig. 1), with a contig N50 of 24.40 Mb.
Chromosome scaffolding and manual curation
Hi-C valid pairs were used to scaffold contigs with 3D-DNA23 (release 180922; run-asm-pipeline.sh -m diploid -r 0), followed by manual curation in Juicebox Assembly Tools v1.11.0824. As an orthogonal check, YAHS v1.2.225 (default parameters) was also run; only concordant joins were retained. Residual gaps were closed with TGS-GapCloser v1.2.126 using both HiFi and ONT reads.
Telomere and centromere features
Telomeric (TTAGGG)n arrays were identified with tidk v0.2.0 (find-telomeres)27 at scaffold ends. Tandem Repeats Finder v4.09.128 annotated satellite-rich regions. Putative centromeres were inferred from (i) long tandem arrays and (ii) local insulation in Hi-C maps using cooltools v0.5.429. Ultimately, we resolved all 24 chromosomes as single contigs and identified 24 centromeres and 20 telomeric tracts (9 upstream, 11 downstream); two chromosomes are capped at both ends, while six lack detectable telomere arrays (Table 1).
Table 1.
Per-chromosome contiguity and coordinates of centromeric and telomeric tracts in the H. variegata assembly.
| Chr | Contigs | Contigs Length (bp) | Gaps | Centromere | Telomere | ||||
|---|---|---|---|---|---|---|---|---|---|
| Start Pos | End Pos | Upstream Start (bp) | Upstream End (bp) | Downstream Start (bp) | Downstream End (bp) | ||||
| LG01 | 1 | 25,034,042 | 0 | 4,614,664 | 5,615,333 | — | — | 25,031,349 | 25,034,042 |
| LG02 | 1 | 24,138,896 | 0 | 17,439,705 | 18,101,404 | 1 | 6018 | — | — |
| LG03 | 1 | 25,875,418 | 0 | 5,441,745 | 6,561,908 | — | — | 25,866,822 | 25,875,418 |
| LG04 | 1 | 29,470,937 | 0 | 24,338,219 | 26,702,069 | 1 | 6968 | 29,468,311 | 29,470,878 |
| LG05 | 1 | 26,971,872 | 0 | 12,994,250 | 13,803,122 | — | — | 26,962,164 | 26,971,872 |
| LG06 | 1 | 36,137,583 | 0 | 16,570,802 | 18,324,963 | — | — | — | — |
| LG07 | 1 | 36,069,280 | 1 | 11,540,992 | 12,045,756 | — | — | 36,061,267 | 36,069,271 |
| LG08 | 1 | 21,907,463 | 0 | 5,135,691 | 5,475,430 | — | — | — | — |
| LG09 | 1 | 23,528,384 | 0 | 15,620,176 | 16,633,023 | 1 | 3448 | — | — |
| LG10 | 1 | 23,324,279 | 0 | 4,017,339 | 5,564,194 | — | — | 23,317,309 | 23,324,195 |
| LG11 | 1 | 23,035,761 | 1 | 7,265,086 | 7,631,375 | 1 | 1596 | — | — |
| LG12 | 1 | 22,421,582 | 0 | 4,554,621 | 5,838,688 | — | — | 22,420,314 | 22,421,582 |
| LG13 | 1 | 28,663,172 | 0 | 18,937,775 | 19,909,336 | — | — | — | — |
| LG14 | 1 | 20,247,748 | 0 | 16,194,167 | 17,061,444 | 1 | 7031 | — | — |
| LG15 | 1 | 21,829,513 | 0 | 6,022,519 | 6,920,656 | 1 | 3444 | 21,824,442 | 21,829,513 |
| LG16 | 1 | 24,398,077 | 0 | 13,451,687 | 14,392,030 | — | — | 24,391,981 | 24,398,077 |
| LG17 | 1 | 21,963,828 | 0 | 21,999,339 | 24,420,426 | — | — | 21,960,072 | 21,963,828 |
| LG18 | 1 | 21,216,350 | 0 | 16,442,062 | 17,086,228 | — | — | — | — |
| LG19 | 1 | 24,020,927 | 0 | 3,399,193 | 4,236,042 | 1 | 9188 | — | — |
| LG20 | 1 | 24,051,685 | 0 | 16,269,960 | 17,924,280 | — | — | — | — |
| LG21 | 1 | 32,735,689 | 0 | 6,006,300 | 7,811,279 | — | — | — | — |
| LG22 | 1 | 24,441,762 | 0 | 16,291,854 | 16,872,664 | 1 | 9647 | — | — |
| LG23 | 1 | 22,760,638 | 0 | 3,688,513 | 4,658,942 | — | — | 22,756,623 | 22,760,638 |
| LG24 | 1 | 26,950,966 | 0 | 20,043,526 | 20,865,722 | 9 | 1697 | — | — |
Repeat annotation
A de novo repeat library was generated with RepeatModeler2 v2.0.430 (including LTRharvest/LTR_retriever v2.9.031 and MITE-Hunter v11-201132), classified with TEclass v2.1.033, merged with Repbase Update34, and applied with RepeatMasker v4.1.635. Tandem repeats were added from TRF v4.10.028; redundant consensus entries were collapsed with cd-hit-est v4.8.1 (-c 0.9)36. In total, 686,404 repeats span 174.00 Mb, accounting for 27.13% of the genome. DNA transposons dominate (14.65%)—notably CACTA (3.70%), ClassII/Unknown (3.31%), Helitron (2.17%), hAT (1.73%) and Tc1–Mariner (1.43%). Retrotransposons comprise 12.45%, mainly LINEs (4.24%) and LTR elements such as Gypsy (2.19%) and LTR/Unknown (3.82%). (Table 2).
Table 2.
Repeat landscape of the H. variegata genome.
| Type | Number | Length | Rate(%) |
|---|---|---|---|
| ClassI:Retroelement | 205,837 | 79,849,924 | 12 |
| ClassI/DIRS | 8,148 | 7,768,409 | 1 |
| ClassI/LINE | 71,968 | 27,158,612 | 4 |
| ClassI/LTR/Copia | 1,777 | 1,021,405 | 0 |
| ClassI/LTR/ERV | 11,069 | 3,535,719 | 1 |
| ClassI/LTR/Gypsy | 27,252 | 14,021,526 | 2 |
| ClassI/LTR/Ngaro | 1,026 | 150,849 | 0 |
| ClassI/LTR/Pao | 749 | 268,739 | 0 |
| ClassI/LTR/Unknown | 74,309 | 24,486,763 | 4 |
| ClassI/SINE | 9,539 | 1,437,902 | 0 |
| ClassII:DNA transposon | 479,723 | 93,944,983 | 15 |
| ClassII/Academ | 1,397 | 347,841 | 0 |
| ClassII/CACTA | 104,647 | 23,714,903 | 4 |
| ClassII/Crypton | 8,202 | 1,802,914 | 0 |
| ClassII/Dada | 2,111 | 427,108 | 0 |
| ClassII/Ginger | 623 | 63,744 | 0 |
| ClassII/Helitron | 110,973 | 13,916,333 | 2 |
| ClassII/IS3EU | 5,604 | 1,148,677 | 0 |
| ClassII/Kolobok | 11,724 | 2,513,427 | 0 |
| ClassII/Maverick | 1,001 | 217,839 | 0 |
| ClassII/Merlin | 2,695 | 368,658 | 0 |
| ClassII/Mutator | 2,686 | 587,678 | 0 |
| ClassII/P | 1,964 | 250,415 | 0 |
| ClassII/PIF-Harbinger | 17,940 | 4,344,538 | 1 |
| ClassII/PiggyBac | 2,471 | 449,735 | 0 |
| ClassII/Sola | 3,876 | 869,822 | 0 |
| ClassII/Tc1-Mariner | 23,962 | 9,166,941 | 1 |
| ClassII/Unknown | 110,581 | 21,256,798 | 3 |
| ClassII/Zator | 2,659 | 533,889 | 0 |
| ClassII/Zisupton | 5,110 | 843,086 | 0 |
| ClassII/hAT | 59,497 | 11,120,637 | 2 |
| Unknown | 832 | 202,241 | 0 |
| srpRNA | 12 | 3,555 | 0 |
| Total | 686,404 | 174,000,703 | 27 |
Gene and transcript annotation
RNA-seq reads (blood only) were QC’d with fastp v0.23.411, aligned with HISAT2 v2.2.137 (--rna-strandness RF -k 10), and assembled with StringTie v2.2.138. Homology-based predictions used GeMoMa v1.939 with teleost references. Ab initio models were predicted by AUGUSTUS v3.5.040. EVidenceModeler v1.1.141 integrated evidence with weights set to RNA-seq = 50, homology = 50, and ab initio = 0.3, prioritizing transcript and homology support while retaining ab initio predictions as low-weight complementary evidence. However, because RNA-seq data were available only from blood, tissue-specific transcripts may be underrepresented; multi-tissue RNA-seq will be incorporated in future annotation updates. Functional annotation used DIAMOND v2.1.942 (NR; --evalue 1e-5 --max-target-seqs1), InterProScan v5.61-93.043, eggNOG-mapper v2.1.1244, and KofamScan v1.3.045. In total, 24,479 protein-coding genes (93%) were functionally annotated in at least one database. Coverage by resource was: TrEMBL 24,065 (91.43%), nr 23,475 (89.18%), Pfam 22,304 (84.74%), GO 21,453 (81.5%), KEGG 21,109 (80.2%), eggNOG 20,981 (79.71%), Swiss-Prot 16,319 (62.0%), and KOG 15,904 (60.42%) (Table 3).
Table 3.
Functional annotation summary of H. variegata protein-coding genes across databases.
| Anno Database | Annotated Number | Annotated Ratio (%) |
|---|---|---|
| GO_Annotation | 21,453 | 81.50 |
| KEGG_Annotation | 21,109 | 80.20 |
| KOG_Annotation | 15,904 | 60.42 |
| Pfam_Annotation | 22,304 | 84.74 |
| Swissprot_Annotation | 16,319 | 62.00 |
| TrEMBL_Annotation | 24,065 | 91.43 |
| eggNOG_Annotation | 20,981 | 79.71 |
| nr_Annotation | 23,475 | 89.18 |
| All_Annotated | 24,479 | 93.00 |
Annotation of non-coding RNA genes
Non-coding RNAs were annotated with tRNAscan-SE v2.0.1246, Infernal v1.1.4 (Rfam 14.9)47 and barrnap v0.9 (https://github.com/tseemann/barrnap). We annotated major ncRNA classes totaling ~4.67 Mb (~0.729% of the genome). The set comprises tRNAs (26,565 copies; 2.00 Mb; 0.3123%), rRNAs (13,293; 2.58 Mb; 0.4026%)—dominated by 5S rRNA (12,706 copies; 1.45 Mb; 0.2266%) with additional 18S (179; 0.0514%) and 28S (232; 0.1204%) and 5.8S (176; 0.0042%)—snRNAs (512; 76.1 kb; 0.01186%) mainly spliceosomal (454; 68.1 kb), and miRNAs (174; 14.1 kb; 0.00221%) (Table 4).
Table 4.
Statistics of the non-coding RNA annotations.
| Type | Copy | Average length (bp) | Total length (bp) | % of genome | |
|---|---|---|---|---|---|
| miRNA | 174 | 81.28 | 14,142 | 0.0022 | |
| tRNA | 26,565 | 75.38 | 2,002,430 | 0.3123 | |
| rRNA | rRNA | 13,293 | 194.23 | 2,581,865 | 0.4026 |
| 18S | 179 | 1840.87 | 329,516 | 0.0514 | |
| 28S | 232 | 3328.81 | 772,285 | 0.1204 | |
| 5S | 12,706 | 114.38 | 1,453,280 | 0.2266 | |
| 5.8S | 176 | 152.18 | 26,784 | 0.0042 | |
| snRNA | 512 | 148.58 | 76,075 | 0.0119 | |
| splicing | 454 | 149.93 | 68,066 | 0.0106 | |
| HACA-box | 25 | 145.32 | 3,633 | 0.0006 | |
| CD-box | 32 | 126.06 | 4,034 | 0.0006 | |
| scaRNA | 1 | 342 | 342 | 0.0001 | |
Data Records
Raw sequencing data are available on the NCBI SRA under BioProject PRJNA130724748. The dataset comprises five experiments: P.variegatus_HiFi (PACBIO_SMRT Revio; SRX30176691), P.variegatus_rnaseq (Illumina NovaSeq X Plus; SRX30158545), P.variegatus_hi-c (Illumina NovaSeq X Plus; SRX30158544), P.variegatus_illumina (Illumina NovaSeq X Plus; SRX30158543), and P.variegatus_ONT (Oxford Nanopore PromethION; SRX30158541).
The primary genome assembly (P. variegatus v1.0) is available at NCBI GenBank under the accession GCA_052674685.149. The GenBank release includes: genomic sequences in fasta format (GCA_052674685.1_ASM5267468v1_genomic.fna.gz) and GenBank flat file format of the genomic sequences in the assembly (GCA_052674685.1_ASM5267468v1_genomic.gbff.gz).
Technical Validation
Read mapping and coverage
Illumina reads were aligned with BWA-MEM2 v2.2.113; PacBio HiFi/ONT reads with minimap2 v2.2619. RNA-seq reads were aligned with HISAT2 v2.2.137. Metrics were computed with samtools v1.2050 and Picard v3.1.1 (https://broadinstitute.github.io/picard/).
QUAST v5.3.051 summarized a 641.26 Mb assembly with a contig N50 of 24.40 Mb. Each of the 24 chromosomes is represented by a single contig; 22 are gap-free, whereas LG07 and LG11 each contain one gap (Table 1). Illumina short reads mapped at 99.81% (proper pairs 95.07%) with a mean depth of 57 × and coverage of 99.93%, 99.79%, 99.56% and 98.59% at ≥ 1 × /5 × /10 × /20 × , respectively. PacBio HiFi reads mapped at 99.97%, with a mean depth of 63× and coverage of 99.98%, 99.73%, 99.49% and 98.19% at ≥ 1 × /5 × /10 × /20 × . Merqury v1.3.152 indicated high k-mer completeness and base-level consensus quality, consistent with the mapping statistics. Gene-space completeness was high by BUSCO v5.6.153 (actinopterygii_odb10; --augustus --long): 96.48% complete (94.97% single-copy; 1.51% duplicated), 0.25% fragmented, and 3.27% missing (3,640 total). Although the BUSCO completeness is high, future re-annotation incorporating multi-tissue transcriptomes and updated pipelines may further improve completeness. CEGMA54 further recovered 457/458 (99.78%) core eukaryotic genes and 246/248 (99.19%) highly conserved CEGs.
KR-normalized Hi-C matrices inspected in Juicebox24 showed 24 chromosome-length interaction blocks with strong on-diagonal signal and minimal off-diagonal noise, consistent with correct scaffolding. Contact probability P(s) curves and insulation profiles computed with cooltools v0.5.429 were smooth and free of abrupt discontinuities, with local insulation minima coinciding with putative centromeric regions, indicating the absence of major structural artifacts.
Acknowledgements
This study was funded by the Leshan Municipal Science and Technology Bureau Key Research Project (Grant No. 23NZD002) and by the Leshan Sub-center of the National Swine Industry Center.
Data availability
All raw sequencing reads (Illumina WGS, Oxford Nanopore, PacBio HiFi, and Hi-C), the final genome assembly (FASTA), and functional/structural annotations (GFF3/FASTA) are available under NCBI BioProject PRJNA130724748. The assembly is deposited at GenBank under accession GCA_052674685.149.
Code availability
All software and versions are listed above. No custom code was used for this study.
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.
Contributor Information
Yong Tang, Email: ty20042028@163.com.
Lingjin Xian, Email: xljin01@163.com.
References
- 1.Li, W., Pu, Y. & Tian, H. Spatial and temporal distribution characteristics and optimum habitat conditions of Paracobitis variegatus in Heishui River. Journal of Fishery Sciences of China30, 515–524 (2023). [Google Scholar]
- 2.Mauice, K. Subspecific differentiation of Paracobit variegatus with comments on its zoogeography. Zoological Research15, 58–67 (1994). [Google Scholar]
- 3.Zhou, Y. Preliminary study on the biology of Paramisgurnus rubripes in the middle reaches of Qingyi River, Sichuan Agricultural University (2007).
- 4.Ma, B. S. et al. Length–weight and length–length relationships of four native fish species from the Yalong River, China. Journal of Applied Ichthyology33, 839–841 (2017). [Google Scholar]
- 5.Guo, Z. Sequencing of mitochondrial genome of Paragonimus rubripes and phylogenetic analysis of Cyprinus carpio, Shaanxi Normal University. (2012).
- 6.Liu, C. Z., Wei, G. H., Hu, J. H. & Liu, X. Y. Complete mitochondrial genome of Paracobitis variegates and its phylogenetic analysis. Mitochondrial DNA Part A27, 2421–2422 (2016).
- 7.Liu, F. et al. The telomere-to-telomere gapless genome of grass carp provides insights for genetic improvement. GigaScience14, giaf059 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nurk, S. et al. The complete sequence of a human genome. Science376, 44–53 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yuan, J. et al. A telomere-to-telomere genome assembly of koi carp (Cyprinus carpio) using long reads and Hi-C technology. GigaScience14, giaf087 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zhang, X., Chen, J., Zhou, W., Wen, J. & Shi, Q. A telomere-to-telomere gap-free genome assembly of the protandrous hermaphrodite Asian seabass (Lates calcarifer). Scientific data12, 1457 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shifu, C. Ultrafast one-pass FASTQ data preprocessing, quality control, and deduplication using fastp. Imeta2, e107 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.De Coster, W. & Rademakers, R. NanoPack2: population-scale evaluation of long-read sequencing data. Bioinformatics39, btad311 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jung, Y. & Han, D. BWA-MEME: BWA-MEM emulated with a machine learning approach. Bioinformatics38, 2404–2413 (2022). [DOI] [PubMed] [Google Scholar]
- 14.Servant, N. et al. HiC-Pro: an optimized and flexible pipeline for Hi-C data processing. Genome biology16, 259 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Durand, N. C. et al. Juicer provides a one-click system for analyzing loop-resolution Hi-C experiments. Cell systems3, 95–98 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Marçais, G. & Kingsford, C. A fast, lock-free approach for efficient parallel counting of occurrences of k-mers. Bioinformatics27, 764–770 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ranallo-Benavidez, T. R., Jaron, K. S. & Schatz, M. C. GenomeScope 2.0 and Smudgeplot for reference-free profiling of polyploid genomes. Nature communications11, 1432 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cheng, H., Concepcion, G. T., Feng, X., Zhang, H. & Li, H. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nature methods18, 170–175 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Li, H. Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics34, 3094–3100 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Guan, D. et al. Identifying and removing haplotypic duplication in primary genome assemblies. Bioinformatics36, 2896–2898 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hu, J., Fan, J., Sun, Z. & Liu, S. NextPolish: a fast and efficient genome polishing tool for long-read assembly. Bioinformatics36, 2253–2255 (2020). [DOI] [PubMed] [Google Scholar]
- 22.Madden, T. The BLAST sequence analysis tool. The NCBI handbook2, 425–436 (2013). [Google Scholar]
- 23.Dudchenko, O. et al. De novo assembly of the Aedes aegypti genome using Hi-C yields chromosome-length scaffolds. Science356, 92–95 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Durand, N. C. et al. Juicebox provides a visualization system for Hi-C contact maps with unlimited zoom. Cell systems3, 99–101 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Zhou, C., McCarthy, S. A. & Durbin, R. YaHS: yet another Hi-C scaffolding tool. Bioinformatics39, btac808 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Xu, M. et al. TGS-GapCloser: a fast and accurate gap closer for large genomes with low coverage of error-prone long reads. GigaScience9, giaa094 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Brown, M. R., de La Rosa, M. G. & Blaxter, M. Tidk: a toolkit to rapidly identify telomeric repeats from genomic datasets. Bioinformatics41, btaf049 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Benson, G. Tandem repeats finder: a program to analyze DNA sequences. Nucleic acids research27, 573–580 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Open2CN Abdennur et al. Cooltools: enabling high-resolution Hi-C analysis in Python. PLOS Computational Biology 20: e1012067 (2024).
- 30.Flynn, J. M. et al. RepeatModeler2 for automated genomic discovery of transposable element families. Proceedings of the National Academy of Sciences117, 9451–9457 (2020). [Google Scholar]
- 31.Ou, S. & Jiang, N. LTR_retriever: a highly accurate and sensitive program for identification of long terminal repeat retrotransposons. Plant physiology176, 1410–1422 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Han, Y. & Wessler, S. R. MITE-Hunter: a program for discovering miniature inverted-repeat transposable elements from genomic sequences. Nucleic acids research38, e199 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Abrusán, G., Grundmann, N., DeMester, L. & Makalowski, W. TEclass—a tool for automated classification of unknown eukaryotic transposable elements. Bioinformatics25, 1329–1330 (2009). [DOI] [PubMed] [Google Scholar]
- 34.Jurka, J. et al. Repbase Update, a database of eukaryotic repetitive elements. Cytogenetic and genome research110, 462–467 (2005). [DOI] [PubMed] [Google Scholar]
- 35.Chen, N. Using Repeat Masker to identify repetitive elements in genomic sequences. Current protocols in bioinformatics5, 4.10. 11–14.10. 14 (2004). [Google Scholar]
- 36.Li, W. & Godzik, A. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics22, 1658–1659 (2006). [DOI] [PubMed] [Google Scholar]
- 37.Kim, D., Paggi, J. M., Park, C., Bennett, C. & Salzberg, S. L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nature biotechnology37, 907–915 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Pertea, M. et al. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nature biotechnology33, 290–295 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Keilwagen, J., Hartung, F. & Grau, J. GeMoMa: homology-based gene prediction utilizing intron position conservation and RNA-seq data. Gene prediction: Methods and protocols Springer, 161–177 (2019).
- 40.Stanke, M., Diekhans, M., Baertsch, R. & Haussler, D. Using native and syntenically mapped cDNA alignments to improve de novo gene finding. Bioinformatics24, 637–644 (2008). [DOI] [PubMed] [Google Scholar]
- 41.Haas, B. J. et al. Automated eukaryotic gene structure annotation using EVidenceModeler and the Program to Assemble Spliced Alignments. Genome biology9, R7 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Buchfink, B., Xie, C. & Huson, D. H. Fast and sensitive protein alignment using DIAMOND. Nature methods12, 59–60 (2015). [DOI] [PubMed] [Google Scholar]
- 43.Jones, P. et al. InterProScan 5: genome-scale protein function classification. Bioinformatics30, 1236–1240 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Cantalapiedra, C. P., Hernández-Plaza, A., Letunic, I., Bork, P. & Huerta-Cepas, J. eggNOG-mapper v2: functional annotation, orthology assignments, and domain prediction at the metagenomic scale. Molecular biology and evolution38, 5825–5829 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Aramaki, T. et al. KofamKOALA: KEGG Ortholog assignment based on profile HMM and adaptive score threshold. Bioinformatics36, 2251–2252 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Chan, P. P., Lin, B. Y., Mak, A. J. & Lowe, T. M. tRNAscan-SE 2.0: improved detection and functional classification of transfer RNA genes. Nucleic acids research49, 9077–9096 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Nawrocki, E. P. & Eddy, S. R. Infernal 1.1: 100-fold faster RNA homology searches. Bioinformatics29, 2933–2935 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.NCBI Sequence Read Archivehttps://identifiers.org/ncbi/insdc.sra:SRP610030 (2025).
- 49.GenBankhttps://identifiers.org/ncbi/insdc.gca:GCA_052674685.1 (2025).
- 50.Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics25, 2078–2079 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Gurevich, A., Saveliev, V., Vyahhi, N. & Tesler, G. QUAST: quality assessment tool for genome assemblies. Bioinformatics29, 1072–1075 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Rhie, A., Walenz, B. P., Koren, S. & Phillippy, A. M. Merqury: reference-free quality, completeness, and phasing assessment for genome assemblies. Genome biology21, 245 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Tegenfeldt, F. et al. OrthoDB and BUSCO update: annotation of orthologs with wider sampling of genomes. Nucleic acids research53, D516–D522 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Parra, G. & Keith Bradnam, I. K. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics23, 1061–1067 (2007). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Citations
- NCBI Sequence Read Archivehttps://identifiers.org/ncbi/insdc.sra:SRP610030 (2025).
- GenBankhttps://identifiers.org/ncbi/insdc.gca:GCA_052674685.1 (2025).
Data Availability Statement
All raw sequencing reads (Illumina WGS, Oxford Nanopore, PacBio HiFi, and Hi-C), the final genome assembly (FASTA), and functional/structural annotations (GFF3/FASTA) are available under NCBI BioProject PRJNA130724748. The assembly is deposited at GenBank under accession GCA_052674685.149.
All software and versions are listed above. No custom code was used for this study.