Comprehensive analysis of 111 Pleuronectiformes mitochondrial genomes: insights into structure, conservation, variation and evolution

Abstract

Background

Pleuronectiformes, also known as flatfish, are important model and economic animals. However, a comprehensive genome survey of their important organelles, mitochondria, has been limited. Therefore, we aim to analyze the genomic structure, codon preference, nucleotide diversity, selective pressure and repeat sequences, as well as reconstruct the phylogenetic relationship using the mitochondrial genomes of 111 flatfish species.

Results

Our analysis revealed a conserved gene content of protein-coding genes and rRNA genes, but varying numbers of tRNA genes and control regions across species. Various gene rearrangements were found in flatfish species, especially for the rearrangement of nad5-nad6-cytb block in Samaridae family, the swapping rearrangement of nad6 and cytb gene in Bothidae family, as well as the control region translocation and tRNA-Gln gene inversion in the subfamily Cynoglossinae, suggesting their unique evolutionary history and/or functional benefit. Codon usage showed obvious biases, with adenine being the most frequent nucleotide at the third codon position. Nucleotide diversity and selective pressure analysis suggested that different protein-coding genes underwent varying degrees of evolutionary pressure, with cytb and cox genes being the most conserved ones. Phylogenetic analysis using both whole mitogenome information and concatenated independently aligned protein-coding genes largely mirrored the taxonomic classification of the species, but showed different phylogeny. The identification of simple sequence repeats and various long repetitive sequences provided additional complexity of genome organization and offered markers for evolutionary studies and breeding practices.

Conclusions

This study represents a significant step forward in our comprehension of the flatfish mitochondrial genomes, providing valuable insights into the structure, conservation and variation within flatfish mitogenomes, with implications for understanding their evolutionary history, functional genomics and fisheries management. Future research can delve deeper into conservation biology, evolutionary biology and functional usages of variations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-025-11204-w.

Background

Fish are one of the most diverse and abundant vertebrate groups on Earth, playing crucial roles in aquatic ecosystems and global fisheries, as well as serving as important model organisms for scientific research [1]. Among them, the Pleuronectiformes, commonly known as flatfish, are particularly intriguing due to their unique morphology, ecological adaptations and economic significance [2]. These fish are characterized by their flattened bodies and asymmetric eyes, with both eyes shifted to one side of the head during the juvenile stage. This distinct morphology allows them to lie on the bottom of aquatic environments and ambush prey. The flatfish encompass a wide range of species distributed globally in marine, freshwater and brackish habitats, from shallow waters to deep-sea environments [3], contributing significantly to developmental and evolutionary studies. In addition, some flatfishes contribute to blue food provision and economic development, including flounders, soles and halibuts [4].

As the primary energy-generating system, mitochondria are essential organelles and found in virtually all eukaryotic cells. More than just powerhouses, they participate in multiple cell signaling cascades and perform a large suite of functions, including but not limited to development, metabolism, aging, disease and immunity [5–7]. Mitochondria originated from a symbiotic relationship between a host prokaryote and an alpha-proteobacterium, therefore, they contain their own genome distinct from nuclear DNA although some genes were transferred to the host’s nuclear genome or lost during evolution [8, 9]. The mitochondrial genome (mitogenome) is a circular DNA molecule with a relatively small size, typically ranging from 15 to 20 kilobases in vertebrate. Due to its maternal inheritance, lack of recombination and relatively fast evolutionary rate, mitogenome has become a powerful tool in phylogenetics and population genetics studies [10]. In particular, complete mitogenome sequences offer a comprehensive view of genome rearrangement, codon usage and genetic diversity, thereby providing valuable insights into the evolutionary history of species. Additionally, analysis on mitogenome facilitates functional studies, and functional mitogenomic variations have been manifested by both laboratory experiments and disease phenotype. For instance, mutations in mitogenomes are associated with metabolic and neurological diseases in human [11]. Using androgenesis rainbow trout (Oncorhynchus mykiss) with identical nuclear backgrounds and different maternal backgrounds, researchers revealed that mitochondrial variation could exert effects on growth rates and oxygen consumption, which may help increase food conversion ratios in aquaculture [12].

The comparative analysis of mitogenomes within the flatfish remains limited. Previous studies have focused primarily on individual genes or small genomic regions, such as cytochrome oxidase subunit I and 16S rRNA [13, 14], for phylogenetic reconstruction and population genetic analyses. However, these approaches often suffer from limited information content and potential homoplasy issues [15, 16]. In contrast, the utilization of complete mitogenome sequences has the potential to overcome these limitations by providing a more holistic and comprehensive dataset. For instance, Shi [17] revealed mitogenome rearrangements, inversions and variations in a limited number of flatfish. Using complete mitogenome sequences of 39 flatfish species and other outgroup species, Campbell investigated the phylogenetic relationships within flatfish and provided weak support for the monophyly of flatfish [18]. Despite these advances, a comprehensive comparative and phylogenetic analysis encompassing a wide range of flatfish species is needed. In particular, repeat analyses are underexplored in flatfish, which would provide valuable resources for structural and functional implications. Such comprehensive analyses would not only fill gaps in our knowledge about the diversity of mitogenome within this fish group but also shed light on evolutionary history and functional significance.

The present study aims to bridge this gap by conducting a comparative and phylogenetic analysis of mitogenomes in 111 flatfish species. First, we seek to characterize the mitogenome features of the selected flatfish species, particularly in terms of gene content, gene rearrangement, codon usage, selective pressure and inter-/intra-specific nucleotide diversity. Second, we intend to construct a robust phylogenetic framework for flatfish based on complete mitogenome sequences, providing new insights into the relationships among flatfish families and genera. Finally, by analyzing tandem repeats and dispersed repeats, we hope to provide valuable resources for better understanding of their functional implications and for potential applications in the selection or breeding of flatfish.

Methods

Mitogenome retrieval

The complete mitogenome sequences of 111 flatfish species from 12 families were retrieved from the NCBI GenBank database (https://www.ncbi.nlm.nih.gov/). The species were chosen to cover a broad taxonomic range within the order Pleuronectiformes. The downloaded files included the gene sequences in FASTA format and their corresponding annotation information in GFF3 format. To confirm the annotation, the re-annotation was conducted by MitoAnnotator [15] and spherical representation was provided.

Mitogenome assembly

We utilized GetOrganelle (v17.7.1) [19] to assemble six mitogenomes of Cynoglossus semilaevis using high-coverage Illumina data from our whole genome sequencing project (unpublished), mainly for intraspecific nucleotide diversity analysis described below. Annotation and spherical representation of generated genomes were performed using MitoAnnotator [15]. All six new mitogenomes and their annotations have been submitted to NCBI (Accession number: PQ665289-PQ665294).

Similarity comparison

To visually compare the mitogenomes of the 111 flatfish species, we used BLAST + [20] and BRIG (BLAST Ring Image Generator) [21]. BRIG generates a circular visualization of multiple genomes, where the similarity between a reference genome and query sequences is shown as concentric rings based on BLAST consistency.

We grouped the 111 species into 6 sets, with Paralichthys olivaceus serving as the reference genome in each set. The first inner ring represented GC content, which is the percentage of guanine (G) and cytosine (C) nucleotides in the genome. The second inner ring depicted GC skew, calculated as the difference in the proportions of G and C in a sliding window, to identify functional regions potentially associated with replication [22] and gene expression regulation [23].

$$\text{GC skew}=\frac{(\text{G}-\text{C})}{(\text{G}+\text{C})}$$

Collinearity analysis

Collinearity analysis was performed using Mauve software [24] to identify conserved and rearranged genomic regions among the 111 flatfish mitogenomes. The species were divided into 11 groups, and the Progressive Mauve algorithm was employed for multiple sequence alignment. Mauve connects similar sequence regions in different genomes with lines, representing conserved blocks. Nonlinear arrangements indicate genomic rearrangements such as insertions, deletions, or inversions.

Codon usage bias analysis

Codon usage bias, the preferential use of specific synonymous codons for the same amino acid, is observed in all domains of life, and every organism possesses a unique codon choice affecting gene expression, translational efficiency, protein folding, protein function and therefore organismal fitness [25–27]. PhyloSuite [28] was employed to extract the coding sequences of the protein-coding genes (PCGs) from the mitogenomes. Subsequently, codon usage indices, including CAI (Codon Adaptation Index), FOP (Frequency of Optimal Codons), CBI (Codon Bias Index), ENC (Effective Number of Codons) and GC3s (GC content at the third codon position), were calculated using CodonW [29]. Using Enc and GC3s, Enc-plot was also conducted to evaluate the degree of preference for imbalanced use of synonymous codons. In addition, Relative Synonymous Codon Usage (RSCU) values were computed to identify codons with biased usage, where RSCU > 1 indicates that the frequency of use of the codon is higher than the average (positive bias), and RSCU < 1 indicates negative bias.

Nucleotide diversity analysis

The CDS sequences of the 13 PCGs were aligned using MAFFT in PhyloSuite, trimmed using Gblocks [30], and concatenated into a single alignment file. Nucleotide diversity (Pi) was analyzed using DnaSP [31], calculated using a sliding window approach with a window size of 200 and a step size of 20. The nucleotide diversity of the control region (CR, also called D-loop region), which is essential for the transcription initiation and the replication of the mitogenome, was analysed using the same pipeline. The calculation formula for Pi value is as follows,

$$\text{Pi}=\frac{2{\sum }_{\text{i}-1}^{\text{n}}{\text{k}}_{\text{i}}-\text{n}}{\text{n}(\text{n}-1)}$$

where N is the number of sequences in the sample, and k_i is the number of different nucleotides at the i^th site in the sequence.

Non-synonymous and synonymous substitutions

Evaluating sequence variations and evolution can be effectively achieved through calculating non-synonymous (Ka) and synonymous (Ks) substitution rates for protein orthologs. MUSCLE-generated [32] protein alignments across species for each PCG were converted into codon comparisons using ParaAT v2.0 [33], which were used for Ka/Ks calculation with KaKs_Calculator v2.0 [34].

Phylogenetic analysis

The complete mitogenome sequences of the 111 flatfish species were aligned using MAFFT in PhyloSuite. PCGs were first translated into amino acids, aligned, and then back-translated into nucleotides. The aligned sequences were optimized using MACSE [35], and trimmed using Gblocks selecting “Codons” as the data type. In contrast, RNA genes (tRNAs and rRNAs) were trimmed by Gblocks using “Nucleotide” as the data type after the alignment with MAFFT. All alignments were then concatenated together. PartitionFinder2 [36] was subsequently used to find the best partitioning strategy and to calculate the best-fit evolutionary models. A maximum likelihood phylogenetic tree was constructed using IQ-TREE [37], with SH-aLRT (Approximate Likelihood Ratio Test) to evaluate the reliability of the tree. The resulting tree was uploaded to iTOL [38], and gene order was added for each mitogenome. In addition, the phylogenetic analysis was also performed based on 13 PCGs, where each of them was aligned separately and concatenated together, followed by the above-mentioned procedure.

Identification of simple sequence repeats (SSRs) and dispersed repeats

SSRs, also called microsatellites, were identified using MISA (MIcroSAtellite identification tool) [39], which scans the genome for repeated nucleotide motifs of different lengths. We set thresholds to detect SSRs of mono-, di-, tri-, tetra-, penta- and hexanucleotide repeats with minimum repeat units of 10, 5, 4, 3, 3 and 3, respectively. A minimum distance of 100 nucleotides was maintained between two SSRs to avoid counting overlapping regions.

Dispersed repeat sequences were analyzed using REPuter [40], which identifies forward, palindromic, reverse and complementary repeats. The minimum repeat size was set to 8 and the Hamming distance was 3 allowing for up to three mismatches. E-value of 0.01 was set as the threshold of significance.

Results

General features of flatfish mitogenomes

The mitogenome length, GC content percentage, NCBI accession number, and available references are listed in Table 1 for each studied species. All mitogenomes contain 13 PCGs (atp6, atp8, cox1, cox2, cox3, cytb, nad1, nad2, nad3, nad4, nad4L, nad5 and nad6) and two rRNA genes (12S rRNA and 16S rRNA), but have different numbers of tRNA genes and CR. Specifically, a total of 108 mitogenomes have 22 tRNA genes, while Citharoides macrolepidotus and Samaris cristatus harbor 23 tRNA genes, and Samariscus latus harbors 24 tRNA genes. The CR was identified in 110 of the 111 mitogenomes, except for Kareius bicoloratus (Fig. S1). In addition, a total of 10 species contain two CRs. Of them, eight were separated by eight tRNA genes and one nd6, including Arnoglossus tenuis, Bothus robinsi, Grammatobothus polyophthalmus, Laeops kitaharae, Laeops lanceolata, Lophonectes gallus, Monolene sessilicauda and Psettina iijimae. Among them, the two CRs of Laeops kitaharae were also observed by other researchers [41]. The two CRs of the other two species, Samaris cristatus and Samariscus latus, are relatively far from each other, separated by 12S rRNA gene, 16S rRNA gene, nd1 gene and several tRNA genes.

Table 1.

Species information and genome characteristics

Family	Scientific name	Length (bp)	GC%	Accession ID	Reference
Paralichthyidae	Paralichthys olivaceus	17,090	46.5	NC_002386	[42]
	Pseudorhombus cinnamoneus	16,599	46.7	NC_022447	N/A
	Paralichthys lethostigma	16,843	45.5	NC_029223	[43]
	Pseudorhombus dupliciocellatus	16,621	47.8	NC_029323	[44]
	Paralichthys dentatus	17,033	44.8	NC_029476	[45]
	Paralichthys adspersus	17,060	46.6	NC_057273	[46]
	Pseudorhombus pentophthalmus	16,684	47.2	NC_065813	[47]
	Gastropsetta frontalis	16,908	46.8	NC_083009	N/A
	Ancylopsetta ommata	16,761	47.6	NC_083030	N/A
	Paralichthys albigutta	17,035	45.0	NC_083031	N/A
	Paralichthys olivaceus x Verasper variegatus	16,946	46.6	NC_082846	N/A
Pleuronectidae	Kareius bicoloratus	15,973	46.7	NC_003176	[48]
	Verasper variegatus	17,273	45.3	NC_007939	[49]
	Verasper moseri	17,588	45.0	NC_008461	[49]
	Hippoglossus hippoglossus	17,546	46.1	NC_009709	[50]
	Hippoglossus stenolepis	17,841	45.7	NC_009710	[50]
	Reinhardtius hippoglossoides	18,017	45.1	NC_009711	[50]
	Platichthys stellatus	17,103	46.8	NC_010966	[51]
	Pleuronichthys cornutus	17,469	43.5	NC_022445	[51]
	Pseudopleuronectes yokohamae	17,383	46.1	NC_028014	[52]
	Cleisthenes herzensteini	17,175	46.0	NC_028021	[53]
	Limanda aspera	17,194	46.1	NC_028281	N/A
	Pleuronichthys japonicus	17,325	43.7	NC_036299	N/A
	Glyptocephalus stelleri	17,142	38.1	NC_060723	[54]
	Pseudopleuronectes herzensteini	16,719	47.0	NC_063673	[55]
	Hippoglossoides dubius	17,216	46.3	NC_066464	N/A
	Acanthopsetta nadeshnyi	17,206	45.8	NC_066465	[56]
	Microstomus achne	16,971	46.0	NC_066466	[57]
	Dexistes rikuzenius	17,494	44.7	NC_066467	[58]
	Parophrys vetulus	16,914	46.2	NC_066930	N/A
	Kareius bicoloratus x Platichthys stellatus	16,874	46.9	NC_080270	N/A
	Platichthys stellatus x Verasper variegatus	16,874	46.9	NC_082285	N/A
	Pseudopleuronectes americanus	17,217	46.5	NC_082555	N/A
	Lepidopsetta bilineata	17,054	45.9	NC_082755	N/A
	Limanda sakhalinensis	17,186	46.5	NC_082768	N/A
	Hippoglossoides robustus	17,215	46.3	NC_082769	N/A
	Myzopsetta proboscidea	17,243	46.1	NC_082770	N/A
	Liopsetta glacialis	16,865	46.8	NC_082783	N/A
	Lyopsetta exilis	17,077	46.4	NC_082803	N/A
	Hippoglossoides elassodon	17,216	46.3	NC_082804	N/A
	Microstomus pacificus	16,898	45.9	NC_082805	N/A
	Psettichthys melanostictus	17,149	46.8	NC_082806	N/A
	Lepidopsetta polyxystra	17,055	46.0	NC_082812	N/A
	Pleuronichthys coenosus	16,863	44.2	NC_083045	N/A
	Eopsetta jordani	17,024	46.5	NC_083049	N/A
	Atheresthes evermanni	17,175	46.0	NC_083172	N/A
	Atheresthes stomias	17,133	44.6	NC_083173	N/A
Soleidae	Solea senegalensis	16,659	45.3	NC_008327	[59]
	Zebrias zebra	16,758	45.2	NC_021377	N/A
	Aesopia cornuta	16,737	44.9	NC_021969	[60]
	Zebrias quagga	17,045	45.1	NC_023225	[61]
	Liachirus melanospilos	17,001	43.6	NC_023539	[62]
	Pseudaesopia japonica	16,789	45.9	NC_023973	[63]
	Pardachirus pavoninus	16,537	45.9	NC_023974	[64]
	Aseraggodes kobensis	16,944	42.5	NC_024285	[65]
	Solea ovata	16,782	45.3	NC_024610	[66]
	Heteromycteris japonicus	17,111	43.7	NC_024921	[67]
	Zebrias zebrinus	16,757	45.2	NC_025199	[68]
	Brachirus orientalis	16,600	43.9	NC_026078	[69]
	Zebrias crossolepis	16,775	45.4	NC_029382	[70]
	Aseraggodes kaianus	16,963	42.8	NC_071940	N/A
Cynoglossidae	Cynoglossus semilaevis	16,731	39.4	NC_012825	[71]
	Cynoglossus abbreviatus	16,417	39.7	NC_014881	[72]
	Paraplagusia japonica	16,694	40.5	NC_021376	[73]
	Cynoglossus sinicus	16,478	39.3	NC_023224	[74]
	Cynoglossus bilineatus	16,454	40.1	NC_023226	[75]
	Paraplagusia bilineata	16,985	39.9	NC_023227	N/A
	Paraplagusia blochii	16,611	39.9	NC_023228	[76]
	Cynoglossus puncticeps	17,142	38.1	NC_023229	[77]
	Cynoglossus itinus	16,915	41.2	NC_023446	N/A
	Cynoglossus gracilis	16,565	38.4	NC_028540	[78]
	Cynoglossus joyneri	16,428	39.6	NC_030256	[79]
	Cynoglossus zanzibarensis	16,569	42.5	NC_030364	[80]
	Cynoglossus senegalensis	16,519	40.1	NC_045034	[81]
	Cynoglossus roulei	16,598	39.2	NC_046735	[82]
	Cynoglossus nanhaiensis	17,130	40.0	NC_050921	[83]
	Symphurus plagiusa	16,771	46.8	NC_083036	N/A
Scophthalmidae	Scophthalmus maximus	17,583	44.3	NC_013183	N/A
	Zeugopterus punctatus	16,800	45.5	NC_052753	N/A
Psettodidae	Psettodes erumei	17,315	46.4	NC_020032	[84]
	Psettodes belcheri	16,747	45.8	NC_083269	[85]
Bothidae	Crossorhombus azureus	16,792	47.4	NC_022446	[51]
	Bothus pantherinus	17,271	51.3	NC_024947	[18]
	Crossorhombus kobensis	16,696	48.3	NC_024949	[18]
	Laeops lanceolata	18,436	49.7	NC_024951	N/A
	Bothus myriaster	16,873	48.2	NC_030365	[86]
	Crossorhombus valderostratus	16,790	48.0	NC_030366	[87]
	Lophonectes gallus	18,642	50.0	NC_030367	[87]
	Chascanopsetta lugubris	17,251	48.9	NC_033392	[87]
	Psettina iijimae	18,080	48.1	NC_044493	[87]
	Arnoglossus tenuis	17,556	46.5	NC_044494	[87]
	Asterorhombus intermedius	16,886	46.3	NC_044725	[88]
	Grammatobothus polyophthalmus	18,170	44.8	NC_045092	[89]
	Laeops kitaharae	18,522	49.8	NC_066468	[41]
	Bothus robinsi	17,673	51.1	NC_083080	N/A
	Monolene sessilicauda	18,156	48.7	NC_083170	N/A
Achiridae	Achirus lineatus	16,577	46.0	NC_023768	[84]
	Trinectes maculatus	16,553	46.5	NC_023769	[84]
	Trinectes inscriptus	16,568	46.4	NC_083156	N/A
Citharidae	Citharoides macrolepidotus	16,497	47.2	NC_024948	[18]
	Lepidoblepharon ophthalmolepis	16,805	46.5	NC_024952	[18]
Cyclopsettidae	Cyclopsetta fimbriata	16,506	47.3	NC_024950	[18]
	Citharichthys cornutus	16,733	41.4	NC_083014	N/A
	Citharichthys arctifrons	16,823	43.4	NC_083038	N/A
	Citharichthys stigmaeus	16,776	44.6	NC_083050	N/A
Rhombosoleidae	Colistium nudipinnis	16,588	45.7	NC_023447	[90]
	Peltorhamphus novaezeelandiae	16,889	44.4	NC_023448	[90]
	Neoachiropsetta milfordi	16,623	48.0	NC_024953	[18]
	Pelotretis flavilatus	16,937	43.9	NC_026284	[90]
Samaridae	Samariscus latus	18,706	42.7	NC_024263	[91]
	Samaris cristatus	18,606	40.0	NC_025903	[84]

The distributions of gene length, GC content and GC skew for all species are shown in Fig. 1. The gene length for each gene is similar across studied species, with very small variations (Fig. 1A). The variation in the cox1 and nad5 was found to be the largest among PCGs in this study, whose gene size was also the largest. This result is consistent with previous findings [92], which showed the close relationship between gene size and length variation among 250 fish species. The lengths of 12S rRNA genes are approximately 950 bp and 16S rRNA genes are around 1700 bp, the wide distribution in the figure is solely because they are plotted together. In contrast, the length variation of CR is very large, which is usually the main cause of variation in mitogenome length. Different genes and CR have varying GC content (Fig. 1B), and most of them are below 50%, consistent with the overall GC percent for mitogenome. The GC percent of CR is the lowest, indicating that CR is AT-rich. Compared with GC content, the GC skew of different genes exhibits more variations (Fig. 1C). The apt8 gene shows the lowest GC skew, and the nad6 gene shows the highest GC skew. The GC skews of PCGs were all negative except for nad6, which was also observed in other fish species [93, 94]. The reason is that nad6 is the only gene on the light strand (L-strand) and all other PCGs are located on the heavy strand (H-strand), demonstrating the strand asymmetry. Generally, the rRNA and tRNA genes exhibit higher GC skew than PCGs, which may be due to their functional roles in protein synthesis and ribosome assembly. The GC and GC skew of tRNA genes exhibit the greatest degree of variability, which may be due to the large number of genes plotted together.

Fig. 1.

The distribution of gene length (a), GC base composition (b) and GC skew (c) in flatfish mitogenomes. CR, control region

Similarity among flatfish mitogenomes

The mitogenomes of the 111 Pleuronectiformes species were compared to assess their similarity. The alignment of 20 representative species (designated Group 1) is shown in Fig. 2, and the other five groups are shown in Fig. S2. The results indicate that mitogenomes exhibited high similarity with most sequences exhibiting over 70% similarity, indicating a high level of homology among the genomes. GC content and GC skew appear randomly distributed throughout the majority of the genome. However, variations and peaks exist, which may have implications for their mitochondrial function and evolution, except for the peaks in the unaligned breaks in the upper area of the image, where biased GC skew only corresponds to the selected reference mitogenome of P. olivaceus.

Fig. 2.

Alignment of mitochondrial genomes among studied representative Pleuronectiformes species. The gap in the circle represents mismatched sequence of genome alignment. GC content (black) and GC skew (purple/green) are at the outermost two circles. Increased GC content and positive GC skew are represented by peaks oriented toward the center of the circle

Genomic rearrangement

To reveal the conservation and variation among mitogenomes, the collinearity analysis was performed (Fig. S3). Collinearity analysis showed strong conservation in gene arrangement of flatfish, with numerous homologous co-linear blocks being observed. There are also many exceptions. Specifically, Samariscus latus and Samaris cristatus exhibited the most significant genomic rearrangements, i.e., the blue-light green block (including nad5-nad6-cytb region) transfer from the right side of the dark red block to the left side. They both belong to the Samaridae family (Table 1), and this family only contained these two species in the study. Moreover, Bothidae family, including 15 species in this study (Table 1), exhibited swapping gene rearrangement between nad6 and cytb (Fig. S1 and S3). These findings indicated that the genomic region near genes of nad5, nad6 and cytb might be a hotspot for rearrangement events during the evolution of flatfish mitogenomes. These results also suggested the unique evolution status of these two families and the functional implication of the genomic rearrangement.

The non-protein coding region exhibits high variations across the flatfish mitogenomes, especially for the green co-linear block on the right side of Fig. S3, indicating highly unconserved structure. In contrast, the purple block in the non-protein coding region only shows in Pleuronichthys japonicus and Pleuronichthys cornutus. Moreover, some large gaps between co-linear blocks in Fig. S3 are observed indicating no homology, and we found that they are CR by examining Fig. S1. The non-protein coding CR is typically between tRNA-pro and tRNA-pre. In addition to the above-mentioned 10 species which contain two CRs, the position of the other 20 species was also not typical (Fig. S1 and S3). The CRs of five mitogenomes are between the PCG cytb and nd6. More specifically, CRs of Bothus myriaster and Chascanopsetta lugubris are between tRNA-Thr and tRNA-Gln, while CRs of Crossorhombus azureus, Crossorhombus kobensis and Crossorhombus valderostratus are between tRNA-Asp and tRNA-Gln. The CRs of subfamily Cynoglossinae, including the Genus Cynoglossus and the Genus Paraplagusia are translocated to the downstream of the 3' end of nd1, more precisely, between nd1 gene and tRNA-Gln gene (Fig. S1), consistent with previous studies [71, 95, 96]. In addition, a tRNA gene inversion was only observed in the subfamily Cynoglossinae, i.e., the tRNA-Gln gene is inverted from the light to the heavy strand, consistent with previous findings [71, 96].

Codon preference of PCGs

To gain insights into the gene expression regulation and evolution, codon usage analysis was performed. The analysis revealed significant biases in the usage of specific codons among the flatfish mitogenomes. The top five frequently used codons are CUA, AUU, CUU, CUC, GCC, corresponding to the amino acids leucine (Leu), isoleucine (Ile), Leu, Leu and alanine (Ala), respectively (Table 2). In terms of RSCU, a total of 31 codons are positively biased, with CGA, CCC, GCC, UCA and CAA being the most significant ones, indicating a strong preference for these codons. Similarly, a set of 28 codons are negatively biased, with GCG, ACG, CCG, UUG, AGU being the most biased ones. The only codon without obvious usage preference is CCU, whose RSCU value is equal to one. The patterns of RSEU are consistent among the analyzed species (Fig. S4), with part of them showing in Fig. 3.

Table 2.

Preference of codon usage in the mitochondrial genomes of Pleuronectiformes

Amino acid	Codon	Times	RSCU	Amino acid	Codon	Times	RSCU
Phe	UUU	12,605	0.96	Tyr	UAU	4811	0.78
	UUC	13,569	1.04		UAC	7509	1.22
Leu	UUA	10,619	0.89	His	CAU	3169	0.54
	UUG	3022	0.25		CAC	8536	1.46
	CUU	16,749	1.4	Gln	CAA	8413	1.61
	CUC	16,407	1.37		CAG	2067	0.39
	CUA	18,471	1.54	Asn	AAU	4353	0.68
	CUG	6678	0.56		AAC	8512	1.32
Ile	AUU	16,952	1.12	Lys	AAA	6747	1.57
	AUC	13,422	0.88		AAG	1824	0.43
Val	GUU	7336	1.15	Asp	GAU	2619	0.61
	GUC	6543	1.03		GAC	5926	1.39
	GUA	8411	1.32	Glu	GAA	7599	1.42
	GUG	3149	0.5		GAG	3093	0.58
Ser	UCU	5403	1.16	Cys	UGU	1124	0.74
	UCC	7140	1.54		UGC	1894	1.26
	UCA	7559	1.63	Trp	UGA	10,237	1.57
	UCG	1591	0.34		UGG	2801	0.43
	AGU	1463	0.31	Arg	CGU	1029	0.48
	AGC	4733	1.02		CGC	1657	0.78
Pro	CCU	6105	1		CGA	4522	2.13
	CCC	10,083	1.65		CGG	1289	0.61
	CCA	6798	1.11	Met	AUA	10,893	1.21
	CCG	1509	0.25		AUG	7090	0.79
Thr	ACU	6578	0.8	Gly	GGU	4438	0.66
	ACC	11,762	1.44		GGC	8586	1.28
	ACA	12,458	1.52		GGA	7980	1.19
	ACG	1927	0.24		GGG	5922	0.88
Ala	GCU	7760	0.82	N/A	UAA	1	4
	GCC	15,428	1.63		UAG	0	0
	GCA	12,741	1.34		AGA	0	0
	GCG	2039	0.21		AGG	0	0

Fig. 3.

Relative synonymous codon usage of protein-coding genes in Pleuronectiformes mitogenomes

Furthermore, a comparative analysis of codon preferences was conducted for each PCG (Table 3), demonstrating that codon bias varied between the genes. In addition, several characteristics were revealed. The frequency of adenine at the third position of synonymous codons (A3s) was the highest in most genes, accounting for 0.422 on average. This was in accordance with the RSCU results, i.e., most codons with RSCU > 1 were end in A (Table 2, Fig. 3). In contrast, guanine at the third position (G3s) had the lowest frequency of 0.144 on average. The GC content at the third position (GC3s) was 0.44, which was roughly equivalent to the overall GC content (0.46). The frequency of optimal codons (FOP) ranged from 0.33 to 0.43. Noting that FOP = 1 indicates that the gene exclusively uses optimal codons, while FOP = 0 suggests that the gene does not use any optimal codons. The codon adaptation index (CAI) value ranged from 0.134 to 0.187, indicating a relatively low degree of match between the codon usage patterns of these genes and those of highly expressed genes. It is noted that a CAI value closer to 1 indicates that the codon usage pattern of a gene is more akin to the preferences observed in highly expressed genes. The range of ENC values is 45.8 to 49.5 for genes in this study, and the overall Enc value is 20 to 61, where 20 indicates that only one codon is used for each amino acid, and 61 indicates that each codon is evenly used. The lower the ENC value, the stronger the preference for codon usage.

Table 3.

Codon preference analysis of 13 protein-coding genes in Pleuronectiformes

Gene	T3s	C3s	A3s	G3s	CAI	CBI	FOP	ENC	GC3s
atp6	0.298	0.364	0.421	0.116	0.134	−0.080	0.33	48.0	0.41
atp8	0.299	0.402	0.494	0.125	0.150	0.016	0.41	47.5	0.40
cox1	0.298	0.383	0.436	0.142	0.174	−0.033	0.40	48.0	0.43
cox2	0.310	0.394	0.480	0.121	0.187	−0.019	0.41	47.4	0.41
cox3	0.270	0.438	0.447	0.101	0.175	0.008	0.43	46.3	0.45
cytb	0.264	0.447	0.421	0.114	0.179	0.021	0.42	47.4	0.48
nad1	0.272	0.392	0.414	0.139	0.154	−0.044	0.37	49.1	0.45
nad2	0.266	0.414	0.401	0.121	0.144	−0.047	0.37	48.5	0.45
nad3	0.280	0.410	0.409	0.135	0.141	−0.055	0.35	46.9	0.45
nad4	0.270	0.385	0.436	0.137	0.137	−0.033	0.38	49.3	0.43
nad4L	0.233	0.399	0.421	0.123	0.148	−0.050	0.37	45.8	0.46
nad5	0.275	0.426	0.425	0.128	0.177	0.002	0.42	49.5	0.46
nad6	0.409	0.122	0.276	0.371	0.155	−0.108	0.33	47.1	0.41
Average	0.288	0.383	0.422	0.144	0.158	−0.032	0.38	47.7	0.44

To investigate the role of mutational pressure in determining the codon usage bias, the ENC-plot analysis was conducted. As shown in Fig. 4, the actual ENC values of all genes are below the expected curve, suggesting that mutation pressure is not the primary factor influencing codon usage and that natural selection plays a significant role [97].

Fig. 4.

ENC-plot of codon usage in mitochondrial genome of Pleuronectiformes. GC3s, GC content at the third codon position; ENC, effective number of codons

Nucleotide diversity of PCGs and CR

Nucleotide diversity analysis of the 13 PCGs revealed varying levels of nucleotide diversity among the genes, ranging from 0.207 to 0.359 (Fig. 5). The atp8 gene exhibited the highest nucleotide diversity (Pi = 0.359), followed by nad6 (Pi = 0.327), nad2 (Pi = 0.323) and atp6 (Pi = 0.306). In contrast, cox1 (Pi = 0.207), cox2 (Pi = 0.249), cox3 (Pi = 0.234) and cytb (Pi = 0.242) were relatively conserved with lower Pi values. The nucleotide diversity of CR was calculated to be 0.282, which was identical to that of nad4L and fell within the mid-range when compared to the Pi values of the 13 PCGs. These findings suggest that they may undergo distinct evolutionary pressures, leading to variations in their nucleotide diversity. Moreover, different regions showed varying levels of nucleotide diversity, providing valuable information for primer designing to distinguish species or conduct evolutionary analysis.

Fig. 5.

Nucleotide diversity in the mitochondrial genomes of Pleuronectiformes. The X axis denotes the nucleotide position, and the Y axis denotes the nucleotide diversity (Pi) value. The continuous red line represents the nucleotide diversity value in different regions of the protein coding gene. The number above each gene represents the average Pi value of the gene

The intraspecific nucleotide diversity was also performed, using data from six mitogenomes of C. semilaevis (Fig. S5) as described in the Methods. As shown in Fig. 6, the intraspecific Pi values range from 0 to 0.006. The low intraspecific nucleotide diversity was consistent with the previous study [98] showing that the nucleotide diversity (Pi) of black carp (Mylopharyngodon piceus) populations ranged from 0.001 to 0.0024. To further confirm the low Pi value, we have also checked the intraspecific Pi value in other animal species [99, 100], which were also consistent with our analysis.

Fig. 6.

Intraspecific nucleotide diversity in the mitochondrial genomes of Cynoglossus semilaevis. The X axis denotes the nucleotide position, and the Y axis denotes the nucleotide diversity (Pi) value. The continuous red line represents the nucleotide diversity value in different regions

Selective pressure of PCGs

To reveal the selective pressure of each mitochondrial PCG in flatfish, the Ka/Ks value was calculated. As shown in Fig. 7, purifying selection exerts great influence on all the PCGs. Specifically, the atp8 gene shows the highest Ka/Ks value, and the cox family genes are generally low with cox1 being the lowest. This pattern is consistent with the result of nucleotide diversity analysis, reinforcing the genetic conservation and evolutionary pressure of each PCG in flatfish mitogenome. Intriguingly, this phenomenon is widely found in most Metazoa [101].

Fig. 7.

Non-synonymous/synonymous ratio (Ka/Ks) of the 13 protein-coding genes in flatfish mitogenomes. The genes (apt6 and atp8) encoding ATPase complex are indicated in red, the genes (cox1–3) encoding cytochrome c oxidase are indicated in green, the gene (cytb) encoding cytochrome b is represented in orange, and the genes (nad1–6 and 4L) encoding NADH dehydrogenase complex are represented in blue

Phylogenetic reconstruction

The phylogenetic analysis of the 111 Pleuronectiformes species revealed a well-resolved phylogenetic tree that largely mirrored the taxonomic classification of the species. The phylogenetic tree using the mitogenome (PCGs and RNA genes) revealed that species were grouped according to their families and genera, with high bootstrap support values indicating robust phylogenetic relationships (Fig. 8). Paralichthyidae and Pleuronectidae were first clustered into one branch, and then formed sister groups with Cyclopsettidae and Bothidae, and clustered with Scophthalmidae, Citharidae, Psettodidae, Rhombosoleidae, Achiridae, Soleidae, Cynoglossidae and Samaridae. Atheresthes stomias and Atheresthes evermanni (highlighted in red arrow) belong to the family Pleuronectidae, but seemed to cluster into one branch with Paralichthyidae. Intriguingly, the family Samaridae (S. latus and S. cristatus) deviated significantly from other species, consistent with the observed large gene rearrangements in these two species.

Fig. 8.

Phylogenetic tree constructed using whole mitogenome (PCGs and RNA genes) of studied Pleuronectiformes. In the left panel, different colors represent different families: light blue, Samaridae; blue, Cynoglossidae; light green, Soleidae; green, Achiridae; pink, Rhombosoleidae; red, Psettodidae; light orange, Citharidae; orange, Scophthalmidae; light purple, Cyclopsettidae; purple, Bothidae; yellow, Paralichthyidae; brown, Pleuronectidae. In the right panel, gene order for each mitogenome is provided, with a color legend displayed in the bottom left corner. The red arrows indicate species which are not grouped into their belonged family

The phylogenetic tree constructed using concatenated separately aligned PCGs showed similar results that clades mostly correspond to the taxonomic classification (Fig. 9), but showing different phylogenetic relationship among families compared with Fig. 8. As shown in Fig. 9, Cynoglossidae was formed sister groups with Soleidae and Samaridae, and then clustered with Achiridae, Psettodidae, Paralichthyidae, Rhombosoleidae, Scophthalmidae, Bothidae, Cyclopsettidae, Paralichthyidae and Pleuronectidae. Specifically, the family Samaridae (S. latus and S. cristatus) does not form a distinct clade away from all other species without the gene rearrangement information. In addition, Symphurus plagiusa (highlighted in blue arrow) of family Cynoglossidae is classified into one clade with the family Samaridae in the tree, but the incongruity is not observed in Fig. 8. These findings provide new insights into the evolutionary history of Pleuronectiformes and support the utility of mitogenome sequences, instead of single gene, for phylogenetic reconstruction.

Fig. 9.

Phylogenetic tree constructed using PCGs of studied Pleuronectiformes mitogenomes. In the left panel, different colors represent different families: light blue, Samaridae; blue, Cynoglossidae; light green, Soleidae; green, Achiridae; pink, Rhombosoleidae; red, Psettodidae; light orange, Citharidae; orange, Scophthalmidae; light purple, Cyclopsettidae; purple, Bothidae; yellow, Paralichthyidae; brown, Pleuronectidae. In the right panel, gene order for each mitogenome is provided, with a color legend displayed in the bottom left corner. The red and blue arrows indicate species which are not grouped into their belonged family. In addition, the blue arrow also indicates the species whose classification is not consistent with that in Fig. 8

SSR and dispersed repeat analysis

The SSRs identified in each species are shown in Table S1. Overall, a total of 227 SSR sites were identified across the studied mitogenomes. Most SSR types were monomeric (43.6%), followed by trimeric (18.9%) and tetrameric (14.1%). The most frequent monomeric SSR was a T repeat monomer, occurring 59 times and making up 59.6% of the total. The most common dimeric and trimeric SSR was AT/TA and TCC. It is noteworthy that mononucleotide A and hexanucleotide repeat were not identified in these mitogenomes. For each species, the number of total SSRs was from zero to seven, with 31 species containing one SSR, 40 species containing 2 SSRs and 21 species containing three SRRs. No SSR was identified in seven species, and Bothus myriaster harbored the most SSR sites, including five monomeric, one trimeric and one tetrameric site.

In addition to simple sequence repeat SSR, we also revealed the presence of forward, palindromic, reverse and complementary repeats in flatfish species. The forward, reverse, palindromic and complementary repeats measured 17–1178, 17–186, 17–41 and 17–38 bp, respectively (Table S2). In general, forward repeats were the most frequent, followed by reverse, palindromic and complementary repeats (Fig. 10 and Table S2). These findings suggest that SSRs and repeat sequences are prevalent in Pleuronectiformes mitogenomes, which may serve as useful molecular markers for evolutionary studies or breeding practices.

Fig. 10.

Numbers of four long repetitive sequences in the mitochondrial genomes of studied Pleuronectiformes. F, R, C and P represent forward, reverse, complementary and palindromic repeats, respectively. Species are indicated by different colors and annotations are displayed at the bottom of the image

Discussion

With the advent of sequencing technology, a large number of mitogenomes have been sequenced, providing valuable resources for evolutionary, structural and functional analyses. The present study provides a comprehensive analysis of 111 mitogenomes of flatfish species, revealed several intriguing aspects of the structure, conservation and variation within these genomes. Our findings not only shed light on the evolutionary trajectories of these important economic and model organisms but also offer valuable insights into their functional genomics.

Our findings reveal both conservation and variation among flatfish mitogenomes, including the gene length, GC skew, codon usage bias, as well as genomic composition and rearrangements. The genomic conservation across flatfish species suggested a shared evolutionary history, and the presence of genomic variation in certain species highlights the potential for lineage-specific evolutionary events. The most striking example of genomic variation is observed in the Samaridae family with the gene count variation and the genomic translocation of nad5-nad6-cytb gene block. For the gene count, Samaris cristatus has 38 genes including 23 tRNA genes, and Samariscus latus harbors 39 genes including 24 tRNA genes, consistent with previous studies [102]. Shtolz also reported that gene counts are consistent among metazoan, with an average of 37 ± 1.4 (SD) genes [103]. Gene rearrangements in animal mitogenomes are usually explained by three main models: the recombination model and the tandem duplication and random loss (TDRL) model, and the tandem duplication and non-random loss (TDNL) model. In S. latus, the model of double replications and random loss has been proposed to account for the rearrangements [91]. The genomic rearrangement could have functional implications, potentially affecting gene expression or protein interactions. For instance, gene rearrangement was found to be correlated with deep-sea adaptation in mussels [104], corals [105] and caridean species [106]. These make sense since mitochondria and mitogenomes play a pivotal role in aerobic respiration to generate energy, and deep sea has unique environmental characteristics including oxygen depletion and limited food availability. Further studies are needed to elucidate the functional consequences of the rearrangement in the mitogenomes of flatfish. In addition, many variations were observed in the non-protein coding regions, indicate a higher degree of evolutionary flexibility in these regions, which may be less constrained by functional constraints, allowing for a more rapid accumulation of mutations and structural changes.

One of the most central and fundamental objectives in biology is the reconstruction of the tree of life [107]. Mitogenomes have been extensively used to infer the tempo and mode of evolutionary changes, as well as the processes underlying speciation and adaptive radiation. The phylogenetic relationships within and among flatfish families are still debated, with different studies proposing conflicting hypotheses [18, 108–110]. In this study, as stated in Methods and Results, the phylogenetic tree was constructed using two different methods, one is based on the complete mitogenomes and the other is based on the concatenated independently aligned PCGs. Both methods revealed that major clades corresponded well to taxonomic families and genera, but the evolutionary relationship among families were different. Notably, using the whole mitogenome information, S. latus and S. cristatus with large genomic arrangement form distinct outgroups (Fig. 8). Using thousands of single-copy orthologous genes, Lv et al. [109] and Hu et al. [111] respectively constructed a phylogenetic tree of nine families of 11 flatfish species and eight families of 10 species, most of the species are the same. In these two trees, Achiridae and Rhombosoleidae formed sister groups, consistent with the phylogenetic tree we have built using whole mitogenome (Fig. 8), but inconsistent with the tree of Fig. 9. In addition, the mode of phylogeny using the PCGs of mitogenomes in five families of 12 flatfish species [112] is more in accordance with the phylogeny shown in Fig. 8. All these previous studies suffer from the limited number of families and species, which may not provide enough evidence for determining which tree in our present study is better. The gene order often remains unchanged over evolutionary, illustrated by that human, mouse, clawed frog and zebrafish all share the same gene arrangement [113–117]. This is largely true in flatfish, but gene order variation and genomic arrangement were identified, and gene order has been believed to convey valuable information in inferring phylogenies [107, 118]. It is worth noting that doubt should be cast on the use of gene order for phylogenetic analysis, since convergent evolution of mitochondrial gene order may occur. While our results point to the importance of whole mitogenome data, more extensive investigations are necessary before definitively concluding which method is more accurate.

The analysis of SSRs and various types of repeat sequences in the flatfish mitogenomes provides valuable insights into the genetic architecture and functional implication of these flatfish. SSRs play roles in DNA replication, transcription, mRNA splicing, translation, gene function and evolution [119–122]. Our comprehensive survey identified a total of 227 SSR sites, with monomeric repeats being the most abundant, followed by trimeric and tetrameric repeats. This distribution pattern suggests a preference for shorter, monomeric repeats within the mitochondrial DNA of flatfish, consistent with the findings in plants [123]. Based on the statistics in FMiR database [124], however, trinucleotide SSR is more dominant than any other SSR types in fish species. The observation emphasizes the unique characteristics of flatfish mitogenomes. The preponderance of T repeat monomers (59.6%) and the absence of (A)n further underscores the bias towards specific nucleotide repetitions in these mitogenomes. In addition to SSRs, the identification of forward, palindromic, reverse and complementary repeats add another layer of complexity to the genetic landscape of flatfish mitogenomes. They may affect replication, recombination and certain functions [40, 125, 126]. For example, a 13-base pair repeat is responsible for large-scale deletion in the human mitogenome, leading to neuromuscular disorders including Kearns-Sayre syndrome and progressive external ophthalmoplegia [126]. However, their specific functions in mitogenome remain largely underexplored in contrast to SSRs, and this disparity in understanding highlights a pivotal avenue for future research endeavors. Moreover, the abundance and diversity of SSRs and dispersed repeats in flatfish mitogenomes underscore their potential utility as molecular markers, aiding in conservation efforts, stock identification and selective breeding practices. SSRs, in particular, have been extensively used in genetic mapping, population genetics and breeding programs due to their high polymorphism and ease of genotyping [127]. Further investigation into the functional roles and population-level variation of these repeats will enhance our understanding of flatfish biology and contribute to the development of effective genetic management strategies.

While our study provides a comprehensive analysis of mitogenomes across 111 flatfish species, several limitations must be acknowledged and further perspectives are proposed. First, more flatfish species and more individual mitogenomes within species should be accommodated to gain a more comprehensive understanding. Second, the use of complete mitogenomes, while powerful, may not capture the full complexity of evolutionary processes occurring at the nuclear level. Studies incorporating both mitochondrial and nuclear markers could provide a more holistic view of flatfish evolution. Third, further investigation should be performed for the species exhibiting unusual gene rearrangements to unravel the mechanisms driving these changes and their potential significance. Fourth, the functional roles of identified repeats remain to be experimentally examined and further used in practice. Future research is warranted to address these questions, paving the way for more comprehensive understanding and utilization.

Conclusions

In this study, using the mitochondrial genomes of 111 flatfish species, we studied the genomic structure, codon preference, nucleotide diversity, selective pressure and repeat sequences, as well as the phylogenetic relationship. The number of protein-coding genes and rRNA genes are conserved, but slight variations exist for tRNA genes and control regions. Various gene rearrangements were identified in flatfish, especially for the family Samaridae and Bothidae, as well as the subfamily Cynoglossinae, suggesting their unique evolutionary history and/or functional implications. Nucleotide diversity and selective pressure analysis indicated the conservation of cytb and cox genes, as well as the conservation levels of specific regions within genes. Phylogenetic analysis using different approaches exhibited different results, highlighting the role of incorporating gene order information in tree reconstruction. Furthermore, the identified repeats provided additional complexity of genome organization and offered valuable markers for evolutionary and functional studies. Our findings represent a significant step towards understanding the genetic architecture, evolutionary dynamics and functional implication of flatfish mitogenomes. The insights gained from this analysis have important implications for the study of adaptive evolution and fisheries management in marine organisms. With the continued advancements in multi-omics technologies, bioinformatics and validation methods, we anticipate that future research will further unravel the evolutionary history, functional implication and underlying mechanisms in these fascinating fish.

Abbreviations

mitogenome

Mitochondrial genome

G

Guanine

C

Cytosine

A

Adenine

T

Thymine

PCG

Protein-coding gene

CAI

Codon Adaptation Index

FOP

Frequency of Optimal Codons

CBI

Codon Bias Index

ENC

Effective Number of Codons

GC3s

GC content at the third codon position of synonymous codons

RSCU

Relative Synonymous Codon Usage

Pi

Nucleotide diversity

CR

Control region

Ka

Non-synonymous substitution rate

Ks

Synonymous substitution rate

SH-aLRT

Approximate Likelihood Ratio Test

SSR

Simple sequence repeat

MISA

MIcroSAtellite identification tool

Leu

Leucine

Ile

Isoleucine

Ala

Alanine

A3s

Adenine at the third position of synonymous codons

G3s

Guanine at the third position of synonymous codons

L-strand

Light strand

H-strand

Heavy strand

TDRL

The tandem duplication and random loss model

TDNL

The tandem duplication and non-random loss model

Authors’ contributions

ST conceived and designed the study. ST and JL performed bioinformatics analysis. ST analyzed and interpreted the results. ST drafted the initial manuscript. WW and ZS commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by the National Key R&D Program of China (Grant No. 2022YFD2400401), Natural Science Foundation of Shandong Province (Grant No. ZR2023QC259), Shandong Key R&D Program for Academician team in Shandong (Grant No. 2023ZLYS02), and Taishan Scholar Youth Project of Shandong Province, China.

Data availability

The datasets analysed during the current study are available in the NCBI repository, with accession numbers being provided in Table 1. Six new mitogenomes of Cynoglossus semilaevis and their annotations have been submitted to NCBI (Accession number: PQ665289-PQ665294).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.