Reference-guided draft genome assembly of the slender walking catfish, Prophagorus nieuhofii

Sample collection, DNA extraction, and genome sequencing

Fish samples were collected during a 2024 expedition aimed at characterizing genetic resources of aquatic animals from a natural population in South Kalimantan, Indonesia (3°21′43.0″S, 114°42′08.3″E). Specimen were captured using bubu traps and held in a pond with 60 cm water depth at 27-28°C for three days to reduce stress. Prior to DNA tissue sampling, fish were anesthetized following¹²: they were placed in a 35-liter bucket with 7 cm of water at 28°C, cooled with liquid ice to 21°C, and then clove oil was added at 160 mg/L. Tissue samples were collected when the fish showed minimal movement after anesthesia. A 10 mg tissue sample was collected from an individual measuring of 32.5 cm in length and weighing 277 g, preserved in DNA shield solution and transported to the laboratory for sequencing. High-quality DNA was extracted using the Quick-DNA high molecular weight (HMW) MagBead kit (Zymo Research) with overnight proteinase K digestion incubation. The DNA extract was quantified using a Qubit fluorometer with an Equalbit 1x ds-DNA HS assay kit for sequencing.

Whole genome sequencing was performed using Oxford Nanopore Technology (ONT) – PromethION. Genomic DNA (1500 ng DNA in 48uL nuclease free water was incubated at 20°C for 30 min, followed by incubation at 65°C for 5 min. Sequencing by ligation was performed using the Ligation Sequencing DNA V14 workflow kit (SQK-LSK114). The basecaller tool was Dorado v0.9.1, using dna_r10.4.1_e8.2_400bps_sup@v5.0.0 basecalling model, with a minimum Q score of 10 and trimming of adapters and barcodes. The quality of the sequencing data was checked using NanoPlot.¹³

Genome assembly and annotation

Genome assembly estimation was done using Flye 2.9.5,¹⁴ while genome scaffolding was conducted with RagTag 2.1.0¹⁵ guided by the reference genome of Clarias gariepinus (GCF_024256425.1). Genome size was estimated using Jellyfish software version 2.3.1¹⁶ and further processed with GenomeScope 2.0 v2.0.1. The assembly statistics were calculated using assembly-stat version 1.0.1. The completeness of the assembly was estimated using Benchmarking Universal Single-Copy Orthologous (BUSCO) version 5.8.2, utilizing miniport.^17–19

Repetitive elements within the genome assembly were identified using RepeatModeler v2.0.6 in conjunction with RepeatMasker v4.1.7 (http://www.repeatmasker.org). Prior to annotation, these repetitive regions were soft masked to minimize interference. Structural genome annotation encompassing gene prediction was conducted using the GALBA pipeline,²⁰ which employs miniprot¹⁷ and AUGUSTUS,²¹ integrating protein data from closely related species as extrinsic evidence. Specifically, protein data from Clarias gariepinus (GCF_024256425.1), Ictalurus furcatus (GCF_023375685.1), Ictalurus punctatus (GCF_001660625.3), and Tachysurus fulvidraco (GCF_022655615.1) were utilized. Functional annotation of the resulting gene predictions was then performed using the ‘funannotate annotate’ command from the Funannotate pipeline (https://funannotate.readthedocs.io/en/latest/install.html), incorporating tools such as InterProScan5,²² Eggnog-Mapper,²³ and SignalP 5.0²⁴ to assign gene names and predict protein functions. Finally, the completeness of the genome annotation was evaluated using BUSCO v5.8.2.¹⁹

Ethical approval

This research was approved by the Ethics Commission for Animal Husbandry and Use, National Research and Innovation Agency (Approval No. 174/KE.02/SK/07/2024). All animal-related procedures were conducted in accordance with institutional guidelines and complied with the ARRIVE 2.0 reporting standards, the checklists for which are available at https://doi.org/10.6084/m9.figshare.29612615.v1.²⁵

Sequence output and genome assembly

Sequencing produced a total of 27,388,841,658 bases from 4,440,560 reads, with 99.8% of bases meeting the designated quality standards. The highest observed mean basecall quality score was 46.4 with a read length of 133, while the longest read reached 6,129,425 with a mean basecall quality score of 14.5 ( Table 1). The draft of genome assembly, as illustrated in the snail graph (https://doi.org/10.5281/zenodo.17429587),²⁶ comprises approximately 5,790 scaffolds, totaling 1.1 gigabases, with the longest scaffold of 54 spanning megabases.

The N50 and N90 values, measuring assembly continuity, are 33.7Mb and 20.6Mb, respectively. The base composition showed 39.5% GC content and 60.5% AT content, whereas the N content (gaps) remained minimal at 0.04%, indicating a highly contiguous and well-assembled genome. Using the Actinopterygii ortholog database, which is based on 3640 universal genes, the assembly demonstrated 98.8% completeness with a low percentage of missing BUSCO (1.15%), suggesting that most expected genes are present. The genome size of this species is similar to that of a related species, Clarias gariepinus, which has a genome size of 969.62 Mb and contig N50 of 33.71 Mb.²⁷ Genome composition based on a 21-mer based characterization shows a heterozygosity rate of 0.78%, while the homozygosity rate was 99.12%.

Genome annotation

Repeats annotation

Repeat annotation analysis revealed that approximately 48.17% of the genome (529,073,132 bp) consisted of repetitive elements ( Table 2). Among these, retroelements accounted for 13.30% of the genome, spanning over 146 million base pairs across 492,154 elements. This category includes SINEs, which comprise 1.93% of the genome, and LINEs, the largest subgroup of retroelements, which occupy 5.33%. The LINEs were mainly composed of L2/CR1/Rex elements (4.01%), followed by the R1/LOA/Jockey, RTE/Bov-B, and L1/CIN4 subfamilies. LTR elements were also prominent, comprising 6.04% of the genome, largely represented by Gypsy/DIRS1 (2.55%) and retroviral elements (1.05%), along with smaller contributions from BEL/Pao and Ty1/Copia. Notably, some retroelement families such as CRE/SLACS were not detected.

DNA transposons represented the largest category of repeats, both in number and genomic coverage, with 924,291 elements occupying 18.56% of the genome (203.9 million base pairs). Within this group, the TC1-IS630-Pogo family was predominant, covering 11.38% of the genome. Other notable contributors included hobo-Activator (3.40%), PiggyBac (0.41%), Tourist/Harbinger (0.64%), and MULE-MuDR (0.01%), while some families, such as En-Spm, showed no representation. Additionally, rolling-circle transposons comprising 30,262 elements and 0.56% of the genome were identified. A substantial portion of the genome (12.13%) contained unclassified elements, amounting to 986,211 entries. These may represent novel, divergent, or currently uncategorized repeat families.

Other repetitive elements included small RNA-related sequences (1.23%), simple repeats (e.g. microsatellites, 3.17%), low-complexity regions (0.31%), and satellite DNA (0.05%). Overall, interspersed repeats alone account for 44.03% of the genome (483.6 million base pairs), underscoring the genomic complexity and abundance of repetitive sequences, especially DNA transposons and retroelements.

Structural and functional annotation

The genome annotation of Prophagorus nieuhofii resulted in the identification of 30,099 protein-coding genes, a count comparable to that of Ictalurus punctatus (approx 31,040 genes), which produced 37,734 predicted transcripts. Relative to the other four catfish species examined ( Table 3), P. nieuhofii exhibits a compact gene architecture and reduced splicing complexity. This is evidenced by the lowest mean number of transcripts per gene at 1.3 (others range from 1.7 to 2.3), and alternative splicing detected in only 18.1% of genes, significantly less than the 32.2% to 48.0% observed in the remaining species. Furthermore, P. nieuhofii has a substantially higher proportion of single-exon genes (12.5%), far exceeding the 3.5% to 4.4% found in the other catfish, which possess a more uniformly multi-exonic architecture. Structural measurements confirm this compactness: P. nieuhofii genes have a smaller average locus length (15,993.4 bp vs. 19,155.7 bp to 22,083.0 bp) and fewer distinct exons per gene (8.9 vs. 11.3 to 12.3). Its mean exon size is also the smallest at 180.2 bp (others range from 269.4 bp to 327.1 bp), resulting in a significantly shorter average transcript size (1,812.4 bp).

In terms of genome composition ( Table 4), the P. nieuhofii genome has the lowest overall content dedicated to coding and genic regions. Exons constitute only 4% of the genome (48 Mb), which is notably low compared to the 8% to 13% seen elsewhere but exhibit the highest GC content at 51% (compared to 45%─46% in the others). Genes collectively occupy 44% of the genome (481 Mb), marking the smallest genic fraction (others range from 55% to 68%). Introns, totaling 233,110, account for 40% of the genome (434 Mb) with an average length of 1,862 bp, also representing the lowest proportional content in the comparison. The structural and compositional differences collectively indicate that the P. nieuhofii genome exhibits a more compact gene structure and less complexity in transcript diversity compared to the other catfish assemblies studied.

To assess the completeness of the annotation, BUSCO analysis was conducted using the actinopterygii_odb10 lineage dataset. The analysis revealed that 97.7% of the 3,640 expected single-copy orthologs were complete, with 79.3% identified as single-copy and 18.5% as duplicated BUSCOs (https://doi.org/10.5281/zenodo.17429587).²⁶ Only 0.7% were fragmented and 1.6% were missing, indicating a highly complete and well-annotated gene set. The high BUSCO score highlights the robustness of the genome annotation, affirming its appropriateness for subsequent biological and comparative analyses.