Skip to main content

Copy number variation and elevated genetic diversity at immune trait loci in Atlantic and Pacific herring



Genome-wide comparisons of populations are widely used to explore the patterns of nucleotide diversity and sequence divergence to provide knowledge on how natural selection and genetic drift affect the genome. In this study we have compared whole-genome sequencing data from Atlantic and Pacific herring, two sister species that diverged about 2 million years ago, to explore the pattern of genetic differentiation between the two species.


The genome comparison of the two species revealed high genome-wide differentiation but with islands of remarkably low genetic differentiation, as measured by an FST analysis. However, the low FST observed in these islands is not caused by low interspecies sequence divergence (dxy) but rather by exceptionally high estimated intraspecies nucleotide diversity (π). These regions of low differentiation and elevated nucleotide diversity, termed high-diversity regions in this study, are not enriched for repeats but are highly enriched for immune-related genes. This enrichment includes genes from both the adaptive immune system, such as immunoglobulin, T-cell receptor and major histocompatibility complex genes, as well as a substantial number of genes with a role in the innate immune system, e.g. novel immune-type receptor, tripartite motif and tumor necrosis factor receptor genes. Analysis of long-read based assemblies from two Atlantic herring individuals revealed extensive copy number variation in these genomic regions, indicating that the elevated intraspecies nucleotide diversities were partially due to the cross-mapping of short reads.


This study demonstrates that copy number variation is a characteristic feature of immune trait loci in herring. Another important implication is that these loci are blind spots in classical genome-wide screens for genetic differentiation using short-read data, not only in herring, likely also in other species harboring qualitatively similar variation at immune trait loci. These loci stood out in this study because of the relatively high genome-wide baseline for FST values between Atlantic and Pacific herring.

Peer Review reports


Genetic screens based on whole-genome sequencing are widely used to identify loci underlying variation in phenotypic traits and ecological adaptation. For instance, this approach has been successfully used to identify hundreds of loci in the Atlantic herring (Clupea harengus) with striking genetic differentiation associated with adaptation to different ecological conditions and timing of reproduction [1]. Genetic screens comparing different species may also be used to identify loci that explain phenotypic differences or have contributed to speciation, if the species are sufficiently closely related and do not show extensive differentiation across the entire genome. This approach has been used to identify loci involved in adaptation in adaptive radiations of closely related species of cichlids [2, 3] and Darwin’s finches [4, 5], among many others. Here we present a genomic comparison of the Atlantic herring and its sister species, the Pacific herring (Clupea pallasi).

Atlantic and Pacific herring are both abundant species with key ecological roles in the North Atlantic and the North Pacific Oceans serving as links between primary plankton production and carnivorous fish, sea birds and sea mammals. The Atlantic herring is a schooling, long-distance migratory fish that is distributed along both the eastern and western shores of the North Atlantic Ocean. Its range extends southward to the English Channel and South Carolina [6, 7], and northeastwards to the White Sea and Barents Sea [8, 9]. In addition, populations of Atlantic herring are adapted to the brackish Baltic Sea. Atlantic herring primarily consists of winter–spring and summer–autumn spawning groups, and each population spawns once per year at a specific time [10, 11]. These populations return to their natal place for reproduction and spawn on rock bottoms [7]. The Pacific herring, in contrast, is distributed along both the eastern and western sides of the Pacific Ocean, ranging from the Chukchi Sea and the Beaufort Sea in the north to Baja California and the Sea of Japan in the south [12, 13]. Populations of Pacific herring also occur in the northeast European seas, including the White Sea, Pechora Sea, and Kara Sea [8, 9]. This species spawns once a year in late winter to spring [11]. Pacific herring also exhibits distinct spawning behaviors compared to Atlantic herring. It spawns in shallow, near-shore habitats on marine vegetation, in contrast with the deeper-spawning Atlantic herring [14]. The two species separated about two million years ago as estimated based on mitochondrial DNA sequence analysis [15]. However, there is a contact zone between the Northeast Atlantic and the European side of the subarctic basin where gene flow occurs [8]. Furthermore, a Pacific/Atlantic hybrid population is present in a subarctic fjord in Norway and has persisted for thousands of years [6].

Our previous whole-genome comparisons of Atlantic herring populations adapted to diverse ecological conditions—primarily regarding spawning season, water salinity at spawning place, and water temperature [10]—revealed striking genetic differentiation at hundreds of loci, with minimal differentiation observed at the remaining loci. This pattern deviates from what is expected under neutral evolution and is consistent with strong signatures of selection [1, 16].

Here, we performed a whole-genome scan for genetic differentiation between Atlantic and Pacific herring, to explore the pattern of genetic differentiation between the two species. However, identification of loci that have contributed to adaptive evolution subsequent to speciation was hampered by high genome-wide sequence divergence. In contrast, the screen turned out to be highly efficient to find loci with remarkably low interspecific genetic differentiation that tend to show high nucleotide diversity both within and between species. We show that these regions, referred to here as high-diversity regions, are highly enriched for genes involved in the immune system and that the remarkably low FST estimates are not due to low interspecies nucleotide divergence but rather to extremely high estimated intraspecies nucleotide diversity. We employ long-read sequencing to decipher genetic diversity at some of these regions.


Negative correlation between interspecies differentiation and intraspecies nucleotide diversity

We performed a genome-wide screening of genetic differentiation between Atlantic and Pacific herring, measured with FST, in 5 kb windows using short-read data from individually sequenced samples (Supplementary Table S1). First, average pairwise genetic differentiation was generally high between Atlantic and Pacific herring (genome average FST = 0.58), most likely explained by the accumulation of nucleotide substitutions and genetic drift subsequent to speciation and limited gene flow between the two species. This pattern is illustrated for chromosome 6 in Fig. 1A (see Supplementary Fig. S1 for all chromosomes). Second, FST exhibits high variation across the chromosome, with most of the heterogeneity in the FST profile explained by variation in the recombination rate. FST values were negatively correlated with recombination rates (r = -0.51, P < 0.001; based on the Atlantic herring recombination map [17]), consistent with findings reported in previous studies [18,19,20]. Stretches of the chromosome with inflated FST values around the midpoint of the chromosome coincide with low recombining regions, and we hypothesize that these are centromeric regions. Third, the high FST baseline is interrupted by intervals of remarkably low FST. For instance, a large interval with distinctly low FST around 6.5 Mb, along with several shorter intervals with the same pattern, is present on chromosome 6 (Fig. 1A). Additional examples of this pattern across all chromosomes are documented in Supplementary Fig. S1.

Fig. 1
figure 1

Distribution of population genetic parameters and repetitive elements calculated in nonoverlapping 5 kb windows in a comparison of Atlantic and Pacific herring. (A) Estimates of genetic differentiation (FST), intraspecies nucleotide diversities in Atlantic (πAtlantic) and Pacific herring (πPacific) and interspecies nucleotide diversity between Atlantic and Pacific herring (dxy) across chromosome 6 are displayed in the first to fourth tracks, respectively. Corresponding plots for all chromosomes are shown in Supplementary Fig. S1. Each dot represents a 5 kb window. Windows in the lower 5th percentile of the genome-wide distribution of FST and the top 5th percentile of the genome-wide distributions of πAtlantic and πPacific are highlighted in red. The bottom track represents the cumulative proportion of repeats across the chromosome. (B) The box plots show the distributions of genetic differentiation (FST) and nucleotide diversities (π and dxy) in the high-diversity regions (red) in comparison with the genomic background (black); the mean values are represented with yellow dots inside the boxes. All comparisons are statistically significant (P < 0.001). (C) The bar chart summarizes the proportion of coverage by different repeat superclasses within the high-diversity regions and background sequences. ***=P < 0.001; NS = not significant

Nucleotide diversity within Atlantic (πAtlantic) and Pacific herring (πPacific) and nucleotide divergence between the two species (dxy) were estimated within 5 kb windows based on short-read data (Fig. 1A and Supplementary Fig. S1). Genome-wide averages of π estimated for the two species were similar, with Pacific herring exhibiting a slightly higher level (πPacific = 0.0028) than Atlantic herring (πAtlantic = 0.0026). The average π estimated in this study is comparable to that in previous reports (πAtlantic = 0.0030; [1]). The genome-wide average for dxy between Atlantic and Pacific herring was estimated at dxy = 0.0053. The U-shaped pattern of the π and dxy plots with low values in the middle of chromosome 6 most likely reflects variation in the recombination rate; the correlations of the recombination rate with πAtlantic, πPacific and dxy were 0.45, 0.24 and 0.24, respectively (P < 0.001 for all comparisons). Further, we note that the regions with near-zero FST are not associated with reduced interspecies sequence divergence, the average dxy for the lower 5th percentile of FST was 0.0064 in comparison with the genome-wide average of 0.0053. The low-FST regions are instead characterized by the very high denominator (intraspecies π) that drastically reduces FST estimates; the average πAtlantic and πPacific for the lower 5th percentile of FST were 0.0058 and 0.0065, in comparison with the genome-wide averages of 0.0026 and 0.0028, respectively.

To further examine these signals and explore the underlying causes of the observed pattern, we extracted windows with FST value lower than the 5th percentile (FST < 0.16) and π values above the 95th percentile in each species (πAtlantic > 0.0073 and πPacific > 0.0067). Based on these criteria, 2,030 windows ( 10 Mb in total) with particularly low FST and high nucleotide diversity in both species compared with the genomic background were identified (Fig. 1B). Throughout the paper, these regions are designated as high-diversity regions. FST has a moderate negative correlation with intraspecies nucleotide diversities estimated for the whole genome (-0.56 with πAtlantic and − 0.42 with πPacific; P < 0.001) but this relationship weakens in the high-diversity regions (-0.21 and − 0.24, respectively; P < 0.001). Estimates of nucleotide diversity within Atlantic and Pacific herring show a strong correlation (0.74; P < 0.001), as expected for closely related species. A strong positive correlation exists between dxy and π in each species (0.77 for πAtlantic and 0.70 for πPacific; P < 0.001); this correlation is even stronger in the high-diversity regions (0.91 and 0.92, respectively; P < 0.001) (Supplementary Fig. S2).

Repetitive elements are not enriched in high-diversity regions

Minimal interspecies differentiation in the high-diversity regions implies that a large proportion of the total variation occurs within species and thus, these signals can potentially be footprints of balancing selection. Balancing selection is characterized by maintaining genetic diversity above neutral expectations, thereby leading to low levels of genetic differentiation between populations in regions under selection [21, 22]. However, an alternative hypothesis is that the high-diversity regions are enriched for repetitive elements. Due to high copy number and sequence identity inherent in repetitive elements, these regions tend to yield unreliable SNP calls because of the difficulty to correctly align short reads to such regions. To assess whether the high-diversity regions were enriched for repeats, we calculated the proportion of each window occupied by different classes of repetitive elements. This analysis showed that the high-diversity regions were not enriched for repetitive elements (Fig. 1A). On the contrary, the average proportion of total repeats observed in the high-diversity regions (0.16) was lower than that in the genomic background (0.22; Welch’s t-test, P < 0.001). More specifically, simple repeats and low complexity repeats were less frequent across the high-diversity regions than in the rest of the genome (0.039 and 0.004 vs. 0.089 and 0.012, respectively; Welch’s t-test, P < 0.001) (Fig. 1C).

High-diversity regions are highly enriched for immune-related genes

We performed gene ontology (GO) term overrepresentation analysis for functional characterization of the genes located in the high-diversity regions. Out of 24,095 protein-coding genes in the Atlantic herring genome, 881 genes had full or partial overlap with the high-diversity regions. We noted that a much higher proportion of genes residing in the high-diversity regions lacked functional annotation compared with genes in the rest of the genome. Specifically, only 474 out of 881 (54%) protein-coding genes in the high-diversity regions had a GO annotation regardless of GO category, in contrast to 17,868 out of 24,095 genes (74%) in the entire genome (χ2 test, P < 0.001). To improve the functional annotation of the herring genes, we employed a pipeline developed by National Bioinformatics Infrastructure Sweden (NBIS) for functional annotation [23]. After the enhancement of the GO term annotations, a total of 21,345 genes were annotated genome-wide with at least one GO term irrespective of GO category, out of which 587 overlapped the high-diversity regions. For biological process (BP) term annotation, 17,783 genes were annotated genome-wide, including 270 genes within the high-diversity regions. Keeping only GO terms associated with a minimum of ten genes, 6,690 BP terms were assessed. The overrepresentation analysis was conducted with Weight01 algorithm implemented in the topGO R package [24], and revealed that the high-diversity regions were highly enriched for genes involved in immune-related processes (Fig. 2). This algorithm accounts for GO dependencies and hierarchical structure [25], but there are nevertheless some degrees of overlap between the top GO terms. These overlapping GO terms are closely related in the GO hierarchy. The interrelationships among the significant terms are illustrated in the GO graph in Supplementary Fig. S3A. The genes labelled with each term, and summary statistics of the GO terms, are compiled in Supplementary Table S2. The results for the overrepresented molecular function (MF) and cellular component (CC) terms are presented in Supplementary Table S2 and Supplementary Fig. S3 (B and C).

Fig. 2
figure 2

Significant biological process terms from GO overrepresentation analysis for genes located in high-diversity regions in Atlantic and Pacific herring. A Bonferroni-corrected significance threshold of P ≤ 0.01 was used. The size of each circle is proportional to the number of enriched genes in the corresponding GO category

The above findings encouraged us to conduct a more detailed examination of some of the high-diversity regions to explore the structure and organization of the genes. For example, a 400 kb region around position 6.5 Mb on chromosome 6 comprises a cluster of genes coding for proteins that contain immunoglobulin-like domains (Fig. 3). These genes are homologous to novel immune-type receptor (NITR) genes in other teleosts [26, 27]. Additional examples of gene organization in high-diversity regions showing a similar pattern are presented in Supplementary Fig. S4, and a comprehensive list of high-diversity regions can be found in Supplementary Table S3. Based on these analyses, we conclude that the most characteristic feature of the high-diversity regions in the comparison of Atlantic and Pacific herring is that they are composed of clusters of immune-related genes.

Fig. 3
figure 3

Zoomed-in plots of a high-diversity region on chromosome 6. The first track shows a marked FST drop in the highlighted region. The second track shows the elevated πAtlantic, πPacific and dxy in the same region; dashed lines depict genome-wide averages. The shaded area in the plots depicts a high-diversity region. Gene organization on the reference assembly is presented in the bottom track (annotation source: NCBI Clupea harengus Annotation Release 102), indicating a cluster of NITR genes within the high-diversity region. Color code for genes is given below the figure

To assess the prevalence of the observed pattern in natural populations of other species, we looked for signals of low differentiation and high nucleotide diversity in previously published data from fish. We identified comparable signals among the radiations of Midas cichlid [2] and in stickleback [28] that contain clusters of immune trait genes similar to those observed in herring (Supplementary Table S4), suggesting that despite being a repeated pattern in teleost fish, such regions have been largely overlooked, or at least left undescribed, in previous comparative genomic studies.

PacBio long-read data reveal extensive copy number variation in high-diversity regions

The analysis of genetic diversity at clusters of closely related genes using short-read data is challenging, partially because the genome assembly may be incorrect with collapsed copy number variation even if the assembly is based on long read data as is the case for the Atlantic herring reference assembly [17]. In addition, errors may occur because of the difficulty to align short reads to the correct copy – in fact, due to structural variation, there is often no “correct copy” for a subset of reads. We therefore generated de novo genome assemblies of two Atlantic herring individuals based on PacBio HiFi long reads to overcome this problem (genome statistics in Table 1). We generated nearly 22 Gb of sequencing data with an average read length of 20 kb for each individual. This allowed us to study copy number variation as well as to more accurately estimate levels of nucleotide diversity. Here we present a detailed analysis of three representative high-diversity regions containing clusters of (i) immunoglobulin heavy variable (IGHV), (ii) CD300e molecule (CLM2), and (iii) interferon-induced protein with tetratricopeptide repeats 10 (IFIT10). These regions are representative of the complex nature of genomic regions harboring immune genes and include multigene/single gene family as well as innate/adaptive immune genes. We also considered the contiguity of assembly contigs and gene annotation when selecting these regions. We first annotated the genes and noted the presence of multiple gene copies along with pseudogenes composed of incomplete or disrupted coding sequences (Table 2). The IGHV locus varied between 30 and 70 gene copies distributed across a region ranging in estimated size from 64 kb to 160 kb on different haplotypes (Fig. 4A).

Table 1 Statistics for haplotype-phased genome assemblies of two Atlantic herring individuals (CS4 and CS5) based on PacBio HiFi sequencing
Table 2 Gene copy numbers of IGHV, CLM2, and IFIT10 in three high-diversity regions in five assembliesa
Fig. 4
figure 4

Genome organization of the IGHV region on the Atlantic herring reference assembly (Chr1:1,882,931-1,976,669) and on four PacBio-based haplotype assemblies. Each vertical black line represents a gene, thus illustrating the copy number variation at the IGHV locus

The Atlantic herring reference assembly [17] contains a cluster of four CLM2 genes and one CLM2 pseudogene on chromosome 22. The PacBio assemblies of this region also contained four full length copies and the number of CLM2 pseudogenes varied from one to four (Fig. 5). Furthermore, the lengths of the PacBio assemblies were all longer than the reference assembly suggesting that duplicated sequences may have been collapsed in the reference assembly. A phylogenetic tree analysis revealed that the CLM2C sequences formed a distinct group whereas the genes designated CLM2A, CLM2B1, and CLM2B2 were all closely related and did not form three distinct allelic series. Thus, it is not possible to align short read sequences to the correct copy (based on genomic location) of CLM2A/B genes. The assembled regions containing a pair of IFIT10 genes on chromosome 23 were all longer on the new PacBio assemblies compared with the reference assembly, and the former contained an IFIT10 pseudogene lacking in the reference assembly (Fig. 6).

Fig. 5
figure 5

Copy number variation of the cluster of CLM2 genes on chromosome 22 in the region 18,433,425–18,434,439 bp of the reference assembly. (A) Genomic organization of CLM2 genes on the Atlantic herring reference assembly and four PacBio-based haplotype assemblies. CLM2A_null lacks exon1 and intron1 which is present in CLM2A. Similarly, CLM2B1_null in CS4_h2 lacks exon1 and intron1 which is present in CLM2B1. CLM2B2_null in CS4_h1 and CS5_h1 has one nonsense mutation in exon2. Sequences referred to as pseudogenes (ψ) contain only fragments of CLM2 coding sequences. (B) Phylogenetic tree of CLM2 coding nucleotide sequences. The color code is the same as in Fig. 5A

Fig. 6
figure 6

Copy number variation of the cluster of IFIT10 genes on chromosome 23 in the region 12,114,407 − 12,116,709 bp of the reference assembly. (A) Genomic organization of IFIT10 genes on five haplotypes. IFIT10B_null in CS4_h2 and CS5_h1 had multiple stop codons throughout the gene length. IFIT10B_null has a deletion of four bp. (B) Phylogenetic tree of IFIT10 coding nucleotide sequences with Allis shad IFIT5 as an outgroup. The color code is the same as in Fig. 6A

We estimated nucleotide diversity (π) for the coding sequences of CLM2 and IFIT10 using short-read and long-read data. For this purpose, we used genomic context to ensure that allelic sequences were compared. This revealed a notable difference in the π values estimated from long-read data compared to those estimated based on short-read data. However, several of these genes had still much higher nucleotide diversity than the genome average of 0.003 (Table 3).

Table 3 Nucleotide diversity (π) based on coding sequences (CDS) of CLM2 and IFIT10 genes


In the present study, a window-based scan of sequence divergence between Atlantic and Pacific herring revealed an unexpected set of low-FST, high-dxy regions that clearly stood out against the genomic background, which is characterized by high FST and moderate dxy. Close inspection of these high-diversity regions unveiled a striking enrichment of immune-related genes, typically forming multigene family clusters. To our knowledge, no previous study has highlighted such regions, but our re-examination of publicly available data revealed a similar enrichment of immune trait genes in regions with low FST between populations and high nucleotide diversity in Midas cichlids [2] and sticklebacks [28]. These findings imply that the observed pattern is widespread, yet has remained largely unexplored as regards to its contribution to genome evolution during speciation and adaptation.

We show that clusters of genes involved in adaptive immunity, such as immunoglobulin, T-cell receptor, and major histocompatibility genes, as well as innate immunity, such as novel immune-type receptor, tripartite motif, tumor necrosis factor receptor, and NOD-like receptor genes are highly enriched at the high-diversity regions detected in this study (Supplementary Table S3). Immune response genes are among the fastest evolving genes and species/lineage-specific clades of different categories of immune genes are commonly observed [30]. This is a reasonable explanation as to why a relatively large proportion of the genes in the high-diversity regions lack functional annotation. Furthermore, the presence of several pseudogenes in the multigene clusters is consistent with these loci following a gene birth-and-death model [31, 32]. Many of the genes present in the high-diversity regions belong to the innate immune system, an observation consistent with the fact that in fish, defense mechanisms against pathogens are skewed toward innate immunity [33, 34], while in mammals, adaptive immune responses are more important for immune protection [35, 36]. Previous studies have also noted lineage/species-specific gene family expansions of genes involved in innate immunity in teleosts relative to other vertebrates, and related these expansions to the evolutionary success of the teleost lineage; this includes genes encoding complement factors [37], toll-like receptors (TLR) [38, 39], novel immune-type receptor (NITR) [26, 40], NACHT-domain and leucine-rich-repeat-containing (NLR) [41, 42], tripartite motif proteins (TRIM) [41, 43], CC chemokines [44] and interferons [45]. Our observations demonstrate that immune-related gene families also underwent rapid and dynamic evolution in the two sister species of herring studied here.

Studying genomic regions harboring multigene family clusters proves to be challenging with the use of short sequence reads because of the difficulty to distinguish allelic and non-allelic sequences with high sequence identities. Thus, population genetic parameters, such as FST, estimated based on short-read sequencing data are prone to be biased at such regions, and hence should be interpreted with caution. We therefore used PacBio long-read data to make it possible to separate haplotypes of homologous chromosomes and identified fine-scale pattern of these complex regions. This analysis revealed copy number variation (CNV), and other re-arrangements, at many of the high-diversity regions. In fact, none of the four haplotypes deduced from two individuals had an identical structure to another haplotype in the three gene regions studied in detail (Figs. 4, 5 and 6). The analysis showed that the extensive genetic diversity at immune-related genes in the Atlantic herring is due to a combination of copy number variation and nucleotide diversity between alleles.

Previous studies have detected a large number of loci contributing to ecological adaptation in Atlantic herring [1, 15]. These studies were facilitated by the extremely low genetic differentiation at neutral loci providing an unusually high signal-to-noise ratio in genetic screens. However, the present study has revealed blind spots in these genetic screens, namely the high-diversity regions described here, that have likely played a critical role in the evolutionary history of these two species. The high nucleotide diversity within populations at these loci will result in low FST estimates also when comparing different populations of Atlantic herring adapted to different environmental conditions, such as those in the marine Atlantic Ocean compared with the brackish Baltic Sea. It is clear that long-read sequencing is required to fully explore the genetic diversity at these loci and how it may contribute to genetic adaptation in herring, and many other vertebrate species.

The Atlantic herring is one of the world’s most abundant vertebrates and a single school may be composed of a billion individuals [46], making them an attractive target for pathogens. The Pacific herring is also an abundant species and its spawning in shallow water may favour spreading of pathogens. It is likely that the genetic diversity at immune-related genes described in this study contributes to the genetic defense against pathogens and thus to overall fitness.


This study has revealed copy number variation and high nucleotide diversity at genes related to both adaptive and innate immune system in Atlantic herring. This is probably a widespread pattern among teleosts. We document that population genetic parameters estimated using short-read sequencing data are unreliable for these regions due to their complexity. They also appear as blind spots in genome scans for regions of genetic differentiation based on FST statistics due to the very high within population nucleotide diversity. We demonstrated that the long-read sequencing approach has the capacity to allow accurate estimation of genetic diversity in these regions and provide new insights into their contribution to ecological adaptation in herring, and in other teleosts.

Materials and methods

Short-read data alignment and variant calling

The Atlantic and Pacific herring samples used in the population genetic analysis were collected in our previous studies [1, 15, 16]. The samples of Atlantic herring were collected from both the Northeast and Northwest Atlantic Ocean, and the Pacific herring samples were captured close to Vancouver. Information about the location and date of sampling, water salinity and spawning season is provided in Supplementary Table S1. Genomic DNA was isolated by standard methods, and sequencing libraries were constructed for each individual. Whole-genome sequencing was carried out with Illumina short-read sequencing and paired-end reads were generated. More detailed procedures for whole-genome sequencing are described in the references provided in Table S1.

We mapped short-reads for each individual to the Atlantic herring reference assembly Ch_v2.0.2 [17] with the BWA-MEM algorithm [47]. Variant calling was performed using DNAseq pipeline in Sentieon suite [48]. We used the Haplotyper algorithm to call variants for each sample, and subsequently used the GVCFtyper algorithm to perform the joint variant calling of all samples. In the variant calling process, all sites, encompassing both variants and invariants, were retained in the output VCF file, denoted as the AllSites VCF. Following the variant calling, indels were excluded, and sites with the following filtering criteria were removed: FS > 60.0, MQRankSum < -12.5, MQ < 40.0, QD < 2.0, ReadPosRankSum < -8.0, and SOR > 3.0. Additionally, genotypes with GQ < 20, DP < 2, and DP > 100 were filtered out. This filtering was performed using the VariantFiltration and SelectVariants tools in GATK [49].

Estimation of population genetic parameters

Population genetic statistics including population differentiation and within and between nucleotide diversity were estimated within nonoverlapping 5 kb windows along each chromosome using pixy [29]. Fixation index (FST) was estimated using Weir and Cockerham’s approach [50]. Average per-site nucleotide differences between all pairs of sequences were calculated within populations (nucleotide diversity, π) and between populations (nucleotide divergence, dxy) within 5 kb windows. Pixy avoids underestimation of nucleotide diversity by using AllSites VCF file. Unbiased estimation of nucleotide diversity is achieved by accounting for missing data in the calculation. Separate files for FST, π and dxy and for different chromosomes were combined and windows with a missing value for any of the parameters were excluded from the final file. Windows with FST values below the 5th percentile and π values above the 95th percentile in each of the two species were extracted for downstream analyses.

Genome screening for repetitive elements

We screened the entire Atlantic herring reference genome [17] for repetitive elements with RepeatMasker [51]. The resulting repetitive elements were classified into superclasses and summarized over the same 5 kb windows that were used for computing diversity parameters. For each window, we calculated the proportion containing different superclasses of repeats.

Gene ontology term overrepresentation analysis

To perform overrepresentation analysis for Gene ontology (GO) terms, we included all protein-coding genes that fully or partially overlapped the windows with FST values falling below the 5th percentile and π values exceeding the 95th percentile in each species. To improve the functional annotation of genes in the herring genome, we used a Nextflow-based pipeline for functional annotation developed by the National Bioinformatics Infrastructure Sweden (NBIS) [23]. This pipeline starts with performing BLAST [52] searches for protein sequences extracted from GFF coordinates against protein database (e-value cutoff was set to 1e-6) and requires UniProt protein fasta file as reference to find the best BLAST matches. Only manually curated proteins (SwissProt proteins) from vertebrates were used. With this approach, it assigns a name to the gene and a description (corresponding to the gene product) to the transcripts. In the next step, it runs InterProScan software package to functionally characterize the genes and assign them functional annotations, including GO terms annotation. The NBIS FunctionalAnnotation pipeline in particular played a role in the improvement of the number of characterized genes with assigned names which increased from 16,499 to 22,579. To further enhance the Gene Ontology (GO) annotation, we obtained orthologs of Atlantic herring genes (with orthology confidence = 1) from BioMart, and assigned the GO annotations of these orthologs, sourced from eleven fish species (zebrafish, stickleback, Atlantic salmon, rainbow trout, Japanese medaka, Asian bonytongue, electric eel, goldfish, Nile tilapia, orange clownfish, common carp) and human, to the corresponding Atlantic herring genes. We then built the final gene-to-GO map file which links each gene identifier with one or more GO terms. This file was used to test the overrepresentation of GO terms.

We used the topGO R package [24] for GO term overrepresentation analysis, which provides the possibility to use a custom Gene-to-GO map. The default algorithm, weight01, was used. This algorithm takes the GO topology into account and tests the significance of each GO term depending on its related terms [25]. The GO hierarchical structure was read into TopGO from GO.db package. Since the analysis was based on gene count, and no gene score was available, Fisher’s exact test was implemented to evaluate the overrepresentation of GO terms. GO terms with less than 10 annotated genes were excluded from the analysis. The number of terms and genes incorporated in the analysis are summarized in Table 4.

Table 4 Numbers of genes and nodes incorporated into the GO term overrepresentation analysis

All distribution plots for population genetic parameters and the proportion of repeats were generated using the ggplot2 R package [53]. Plots of gene organization in the high-diversity regions on the reference assembly were created using the gggenes R package [54], and GO graphs were visualized with the topGO R package [24].

Generation of PacBio long-read data

We generated PacBio HiFi sequencing data from two individuals of Atlantic herring, CS4 and CS5, captured on November 25, 2019 in the Celtic Sea. Testis samples were collected and flash-frozen on-site. DNA was extracted using a Circulomics Nanobind Tissue Big DNA Kit (NB-900-701-001) and sequenced using Pacific Biosciences (PacBio) High Fidelity (HiFi) technology. HiFi reads are known to have high accuracy (above 99.8%) and long contiguity (average read length of 13.5 kb) [55] hence we used them to build haploid de novo genome assemblies using hifiasm (v0.16.1-r375) [56]. The four haplotypes from the PacBio assemblies from the two individuals along with the sequences from the reference assembly were used for further analysis.

We selected three regions from the genome scan to characterize the high-diversity regions. As the GO analysis showed a strong enrichment of immune response genes in the high-diversity regions, we selected three representative regions that illustrate the complex nature of immune genes including multigene/single gene family and innate/adaptive immune genes. We also considered the contiguity of assembly contigs and gene annotation for selecting these regions. One of the regions encoded for immunoglobulin heavy variable (IGHV) multigene family on chromosome 1, and the other two regions were annotated as containing single genes and without an indication of a multigene family on chromosomes 22 and 23, with Ensembl IDs as ENSCHAG00000003891 and ENSCHAG00000015470, respectively. The latter two genes lacked gene name in the Ensembl database, hence to name them, we used nucleotide BLAST [52] to find the most similar gene sequence. CMRF35-like molecule 2 (CLM2) and interferon-induced protein with tetratricopeptides 10 (IFIT10) were found to be highly similar to ENSCHAG00000003891 and to ENSCHAG00000015470, respectively.

To check for additional copies of these genes, we used nucleotide BLAST [52] and MUMmer [57] aligner using the coding sequences. It resulted in total 40 IGHV, four CLM2, and two IFIT10 genes, and a few pseudogenes. Additional IGHV genes lacked Ensembl IDs, hence these were manually annotated. Additional CLM2 genes were ENSCHAG00000003799, ENSCHAG00000003851, and ENSCHAG00000003927, while additional IFIT10 was ENSCHAG00000015478. We named the copies of these additional genes in the order they occurred on the genome – CLM2A, CLM2B1, CLM2B2, and CLM2C on Chr22, and IFIT10A and IFIT10B on Chr23.

To annotate these genes on the PacBio assemblies, we used LiftOff [58] with additional “-polish and -copies” parameters. Because IGHV is a multigene family, LiftOff algorithm was not successful in annotating all IGHV genes, hence we performed manual curation and used nucleotide BLAST [52] to find homologous sequences using manually annotated reference sequences as queries.

We used gggenomes R package [59] to visualize genomic organization of the annotated regions. The phylogenetic tree was constructed using ape R package [60] and visualized using ggtree [61]. Homologous genes from the clupeid species Alosa alosa (allis shad) were used as an outgroup to root the tree. Genes that were null due to incomplete sequence lengths were excluded for the construction of phylogenetic tree.

To assess the reliability of π estimates for the high-diversity regions derived from short-read data, we calculated π for the coding sequences of CLM2 and IFIT10 genes, once based on short-read data and once using long-read data. To calculate π based on short-read data, we utilized pixy [29] with the AllSites VCF file, keeping only CS4 and CS5 samples in the calculation. For accurate estimation of π from long-read data, coding sequences only from untruncated gene copies were extracted from four haploid PacBio assemblies obtained from CS4 and CS5. The sequences were subsequently aligned using the msa R package [62], and π was calculated using the pegas R package [63].

Data availability

The raw long-read sequencing data used in this study are deposited in the NCBI Sequence Read Archive (SRA) under the BioProject accession number PRJNA1023520.


  1. Han F, Jamsandekar M, Pettersson ME, Su L, Fuentes-Pardo AP, Davis BW, et al. Ecological adaptation in Atlantic herring is associated with large shifts in allele frequencies at hundreds of loci. eLife. 2020;9:e61076.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Kautt AF, Kratochwil CF, Nater A, Machado-Schiaffino G, Olave M, Henning F, et al. Contrasting signatures of genomic divergence during sympatric speciation. Nature. 2020;588:106–11.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. McGee MD, Borstein SR, Meier JI, Marques DA, Mwaiko S, Taabu A, et al. The ecological and genomic basis of explosive adaptive radiation. Nature. 2020;586:75–9.

    Article  CAS  PubMed  Google Scholar 

  4. Rubin C-J, Enbody ED, Dobreva MP, Abzhanov A, Davis BW, Lamichhaney S, et al. Rapid adaptive radiation of Darwin’s finches depends on ancestral genetic modules. Sci Adv. 2022;8:eabm5982.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Lamichhaney S, Berglund J, Almén MS, Maqbool K, Grabherr M, Martinez-Barrio A, et al. Evolution of Darwin’s finches and their beaks revealed by genome sequencing. Nature. 2015;518:371–5.

    Article  CAS  PubMed  Google Scholar 

  6. Pettersson ME, Fuentes-Pardo AP, Rochus CM, Enbody ED, Bi H, Väinölä R, et al. A long-standing hybrid population between Pacific and Atlantic herring in a subarctic fjord of Norway. Genome Biol Evol. 2023;15:evad069.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Bekkevold D, Berg F, Polte P, Bartolino V, Ojaveer H, Mosegaard H, et al. Mixed-stock analysis of Atlantic herring (Clupea harengus): a tool for identifying management units and complex migration dynamics. ICES J Mar Sci. 2023;80:173–84.

    Article  Google Scholar 

  8. Semenova AV, Stroganov AN. Introgressive hybridization between the Atlantic and Pacific herring (Clupea harengus and Clupea pallasii) in the White Sea, Barents and Kara Seas evidenced by microsatellites. Conserv Genet. 2018;19:143–53.

    Article  Google Scholar 

  9. Laakkonen HM, Strelkov P, Lajus DL, Väinölä R. Introgressive hybridization between the Atlantic and Pacific herrings (Clupea harengus and C. Pallasii) in the north of Europe. Mar Biol. 2015;162:39–54.

    Article  CAS  Google Scholar 

  10. Andersson L, André C, Johannesson K, Pettersson M. Ecological adaptation in Cod and herring and possible consequences of future climate change in the Baltic Sea. Front Mar Sci. 2023;10:1101855.

  11. Haegele CW, Schweigert JF. Distribution and characteristics of herring spawning grounds and description of spawning behavior. Can J Fish Aquat Sci. 1985;42:s39–55.

    Article  Google Scholar 

  12. Orlova SYu, Rastorguev S, Bagno T, Kurnosov D, Nedoluzhko A. Genetic structure of marine and lake forms of Pacific herring Clupea pallasii. PeerJ. 2021;9:e12444.

    Article  Google Scholar 

  13. Thompson SA, Sydeman WJ, Thayer JA, Weinstein A, Krieger KL, St M et al. Trends in the Pacific herring (Clupea pallasii) metapopulation in the California Current ecosystem. 2017;58:77–94.

  14. Hay DE, McCarter PB, Daniel KS, Schweigert JF. Spatial diversity of Pacific herring (Clupea pallasi) spawning areas. ICES J Mar Sci. 2009;66:1662–6.

    Article  Google Scholar 

  15. Martinez Barrio A, Lamichhaney S, Fan G, Rafati N, Pettersson M, Zhang H, et al. The genetic basis for ecological adaptation of the Atlantic herring revealed by genome sequencing. eLife. 2016;5:e12081.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Lamichhaney S, Fuentes-Pardo AP, Rafati N, Ryman N, McCracken GR, Bourne C, et al. Parallel adaptive evolution of geographically distant herring populations on both sides of the North Atlantic Ocean. Proc Natl Acad Sci. 2017;114:E3452–61.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Pettersson ME, Rochus CM, Han F, Chen J, Hill J, Wallerman O, et al. A chromosome-level assembly of the Atlantic herring genome—detection of a supergene and other signals of selection. Genome Res. 2019;29:1919–28.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Wang L, Liu S, Yang Y, Meng Z, Zhuang Z. Linked selection, differential introgression and recombination rate variation promote heterogeneous divergence in a pair of yellow croakers. Mol Ecol. 2022;31:5729–44.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Keinan A, Reich D. Human population differentiation is strongly correlated with local recombination rate. PLOS Genet. 2010;6:e1000886.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Nachman MW, Payseur BA. Recombination rate variation and speciation: theoretical predictions and empirical results from rabbits and mice. Philos Trans Biol Sci. 2012;367:409–21.

    Article  Google Scholar 

  21. Brandt DYC, César J, Goudet J, Meyer D. The effect of balancing selection on population differentiation: a study with HLA genes. G3 (Bethesda). 2018;8:2805–15.

  22. Fischer MC, Foll M, Heckel G, Excoffier L. Continental-scale footprint of balancing and positive selection in a small rodent (Microtus arvalis). PLoS ONE. 2014;9:e112332.

    Article  PubMed  PubMed Central  Google Scholar 

  23. NBIS functional annotation pipeline. Accessed 20 Dec 2023.

  24. Alexa A, Rahnenfuhrer J. topGO: enrichment analysis for gene ontology. Accessed 22 Dec 2023.

  25. Alexa A, Rahnenführer J, Lengauer T. Improved scoring of functional groups from gene expression data by decorrelating GO graph structure. Bioinformatics. 2006;22:1600–7.

    Article  CAS  PubMed  Google Scholar 

  26. Desai S, Heffelfinger AK, Orcutt TM, Litman GW, Yoder JA. The medaka novel immune-type receptor (NITR) gene clusters reveal an extraordinary degree of divergence in variable domains. BMC Evol Biol. 2008;8:177.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Yoder JA, Mueller MG, Wei S, Corliss BC, Prather DM, Willis T, et al. Immune-type receptor genes in zebrafish share genetic and functional properties with genes encoded by the mammalian leukocyte receptor cluster. Proc Natl Acad Sci U S A. 2001;98:6771–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Yamasaki YY, Kakioka R, Takahashi H, Toyoda A, Nagano AJ, Machida Y, et al. Genome-wide patterns of divergence and introgression after secondary contact between Pungitius sticklebacks. Philos Trans R Soc B Biol Sci. 2020;375:20190548.

    Article  Google Scholar 

  29. Korunes KL, Samuk K. Pixy: unbiased estimation of nucleotide diversity and divergence in the presence of missing data. Mol Ecol Resour. 2021;21:1359–68.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Hill T, Koseva BS, Unckless RL. The genome of Drosophila innubila reveals lineage-specific patterns of selection in immune genes. Mol Biol Evol. 2019;36:1405–17.

  31. Nei M, Rooney AP. Concerted and birth-and-death evolution of multigene families. Annu Rev Genet. 2005;39:121–52.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Nei M, Gu X, Sitnikova T. Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc Natl Acad Sci. 1997;94:7799–806.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Uribe C, Folch H, Enriquez R, Moran G. Innate and adaptive immunity in teleost fish: a review. Veterinární medicína. 2011;56:486–503.

    Article  CAS  Google Scholar 

  34. Magnadóttir B. Innate immunity of fish (overview). Fish Shellfish Immunol. 2006;20:137–51.

    Article  PubMed  Google Scholar 

  35. Han Y, Gao H, Xu J, Luo J, Han B, Bao J, et al. Innate and adaptive immune responses against microsporidia infection in mammals. Front Microbiol. 2020;11:1468.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Litman GW, Rast JP, Fugmann SD. The origins of vertebrate adaptive immunity. Nat Rev Immunol. 2010;10:543–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Najafpour B, Cardoso JCR, Canário AVM, Power DM. Specific evolution and gene family expansion of complement 3 and regulatory factor H in fish. Front Immunol. 2020;11:568631.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Solbakken MH, Tørresen OK, Nederbragt AJ, Seppola M, Gregers TF, Jakobsen KS, et al. Evolutionary redesign of the Atlantic cod (Gadus morhua L) toll-like receptor repertoire by gene losses and expansions. Sci Rep. 2016;6:25211.

  39. Star B, Nederbragt AJ, Jentoft S, Grimholt U, Malmstrøm M, Gregers TF, et al. The genome sequence of Atlantic cod reveals a unique immune system. Nature. 2011;477:207–10.

  40. Ferraresso S, Kuhl H, Milan M, Ritchie DW, Secombes CJ, Reinhardt R, et al. Identification and characterisation of a novel immune-type receptor (NITR) gene cluster in the European sea bass, Dicentrarchus labrax, reveals recurrent gene expansion and diversification by positive selection. Immunogenetics. 2009;61:773–88.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Suurväli J, Garroway CJ, Boudinot P. Recurrent expansions of B30.2-associated immune receptor families in fish. Immunogenetics. 2022;74:129–47.

    Article  PubMed  Google Scholar 

  42. Howe K, Schiffer PH, Zielinski J, Wiehe T, Laird GK, Marioni JC, et al. Structure and evolutionary history of a large family of NLR proteins in the zebrafish. Open Biol. 2016;6:160009.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Boudinot P, van der Aa LM, Jouneau L, Du Pasquier L, Pontarotti P, Briolat V, et al. Origin and evolution of TRIM proteins: new insights from the complete TRIM repertoire of zebrafish and pufferfish. PLoS ONE. 2011;6:e22022.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Liu F, Wang T, Hu Y, Tian G, Secombes CJ, Wang T. Expansion of fish CCL20_like chemokines by genome and local gene duplication: Characterisation and expression analysis of 10 CCL20_like chemokines in rainbow trout (Oncorhynchus mykiss). Dev Comp Immunol. 2020;103:103502.

    Article  CAS  PubMed  Google Scholar 

  45. Liu F, Wang T, Petit J, Forlenza M, Chen X, Chen L, et al. Evolution of IFN subgroups in bony fish – 2. Analysis of subgroup appearance and expansion in teleost fish with a focus on salmonids. Fish Shellfish Immunol. 2020;98:564–73.

    Article  CAS  PubMed  Google Scholar 

  46. Parrish JK, Viscido SV, Grünbaum D. Self-organized fish schools: an examination of emergent properties. Biol Bull. 2002;202:296–305.

    Article  PubMed  Google Scholar 

  47. Li H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv. 2013;1303.3997v2.

  48. Freed D, Aldana R, Weber JA, Edwards JS. The Sentieon Genomics Tools - A fast and accurate solution to variant calling from next-generation sequence data. BioRxiv. 2017;115717.

  49. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The genome analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20:1297–303.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Weir BS, Cockerham CC. Estimating F-statistics for the analysis of population structure. Evolution. 1984;38:1358–70.

    CAS  PubMed  Google Scholar 

  51. Smit AFA, Hubley R, Green P. RepeatMasker Open-4.0. 2013–2015. Accessed 22 Dec 2023.

  52. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215:403–10.

    Article  CAS  PubMed  Google Scholar 

  53. Wickham H. ggplot2: elegant graphics for data analysis. Springer-Verlag New York. 2016. Accessed 22 Dec 2023.

  54. Wilkins D. gggenes: draw gene arrow maps in ggplot2. 2023. Accessed 22 Dec 2023.

  55. Wenger AM, Peluso P, Rowell WJ, Chang P-C, Hall RJ, Concepcion GT, et al. Accurate circular consensus long-read sequencing improves variant detection and assembly of a human genome. Nat Biotechnol. 2019;37:1155–62.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  56. Cheng H, Concepcion GT, Feng X, Zhang H, Li H. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nat Methods. 2021;18:170–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Marçais G, Delcher AL, Phillippy AM, Coston R, Salzberg SL, Zimin A. MUMmer4: a fast and versatile genome alignment system. PLOS Comput Biol. 2018;14:e1005944.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Shumate A, Salzberg SL. Liftoff: accurate mapping of gene annotations. Bioinformatics. 2021;37:1639–43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Hackl T, Ankenbrand M, van Adrichem B. gggenomes: a grammar of graphics for comparative genomics. 2023. Accessed 20 Dec 2023.

  60. Paradis E, Claude J, Strimmer K. APE: analyses of phylogenetics and evolution in R language. Bioinformatics. 2004;20:289–90.

    Article  CAS  PubMed  Google Scholar 

  61. Yu G, Smith DK, Zhu H, Guan Y, Lam TT-Y. Ggtree: an r package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods Ecol Evol. 2017;8:28–36.

    Article  Google Scholar 

  62. Bodenhofer U, Bonatesta E, Horejš-Kainrath C, Hochreiter S. Msa: an R package for multiple sequence alignment. Bioinformatics. 2015;31:3997–9.

    Article  CAS  PubMed  Google Scholar 

  63. Paradis E. Pegas: an R package for population genetics with an integrated-modular approach. Bioinformatics. 2010;26:419–20.

    Article  CAS  PubMed  Google Scholar 

Download references


We extend our gratitude to our colleagues and former lab members for their assistance in collecting herring samples and providing technical support throughout laboratory procedures and sequencing. Support from NBIS (National Bioinformatics Infrastructure Sweden) is sincerely acknowledged. The National Genomics Infrastructure (NGI)/Uppsala Genome Center provided service in massive parallel sequencing and the computational infrastructure was provided by the Swedish National Infrastructure for Computing (SNIC) at UPPMAX partially funded by the Swedish Research Council (2018-05973).


The project was financially supported by Vetenskapsrådet (2017–02907; to LA) and Knut and Alice Wallenberg Foundation (KAW 2016.0361; to LA).

Open access funding provided by Uppsala University.

Author information

Authors and Affiliations



LA conceived and designed the study. FMS, MJ, and EE conducted the bioinformatic analysis. MP contributed to the bioinformatic analysis. FMS, MJ, and LA wrote the manuscript with input from other authors. All authors reviewed and approved the final manuscript.

Corresponding author

Correspondence to Leif Andersson.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Mohamadnejad Sangdehi, F., Jamsandekar, M.S., Enbody, E.D. et al. Copy number variation and elevated genetic diversity at immune trait loci in Atlantic and Pacific herring. BMC Genomics 25, 459 (2024).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: