Background Understanding genetic architecture is essential for determining how traits will change in response to evolutionary processes such as selection, genetic drift and/or gene flow. In Atlantic salmon, age at maturity is an important life history trait that affects factors such as survival, reproductive success, and growth. Furthermore, age at maturity can seriously impact aquaculture production. Therefore, characterizing the genetic architecture that underlies variation in age at maturity is of key interest. Results Here, we refine our understanding of the genetic architecture for age at maturity of male Atlantic salmon using a genome-wide association study of 11,166 males from a single aquaculture strain, using imputed genotypes at 512,397 single nucleotide polymorphisms (SNPs). All individuals were genotyped with a 50K SNP array and imputed to higher density using parents genotyped with a 930K SNP array and pedigree information. We found significant association signals on 28 of 29 chromosomes (P-values: 8.7 × 10−133–9.8 × 10−8), including two very strong signals spanning the six6 and vgll3 gene regions on chromosomes 9 and 25, respectively. Furthermore, we identified 116 independent signals that tagged 120 candidate genes with varying effect sizes. Five of the candidate genes found here were previously associated with age at maturity in other vertebrates, including humans. Discussion These results reveal a mixed architecture of large-effect loci and a polygenic component that consists of multiple smaller-effect loci, suggesting a more complex genetic architecture of Atlantic salmon age at maturity than previously thought. This more complex architecture will have implications for selection on this key trait in aquaculture and for management of wild salmon populations.

Infectious pancreatic necrosis virus (IPNV) is the cause of one of the most prevalent diseases in farmed Atlantic salmon (Salmo salar). A quantitative trait locus (QTL) has been found to be responsible for most of the genetic variation in resistance to the virus. Here we describe how a linkage disequilibrium-based test for deducing the QTL allele was developed, and how it was used to produce IPN-resistant salmon, leading to a 75% decrease in the number of IPN outbreaks in the salmon farming industry. Furthermore, we describe how whole-genome sequencing of individuals with deduced QTL genotypes was used to map the QTL down to a region containing an epithelial cadherin (cdh1) gene. In a coimmunoprecipitation assay, the Cdh1 protein was found to bind to IPNV virions, strongly indicating that the protein is part of the machinery used by the virus for internalization. Immunofluorescence revealed that the virus colocalizes with IPNV in the endosomes of homozygous susceptible individuals but not in the endosomes of homozygous resistant individuals. A putative causal single nucleotide polymorphism was found within the full-length cdh1 gene, in phase with the QTL in all observed haplotypes except one; the absence of a single, all-explaining DNA polymorphism indicates that an additional causative polymorphism may contribute to the observed QTL genotype patterns. Cdh1 has earlier been shown to be necessary for the internalization of certain bacteria and fungi, but this is the first time the protein is implicated in internalization of a virus.

Resistance against speci¢c diseases a¡ecting aquaculture species often show moderate to high herit-

Reliability of genomic selection (GS) models was tested in an admixed population of Atlantic salmon, originating from crossing of several wild subpopulations. The models included ordinary genomic BLUP models (GBLUP), using genome-wide SNP markers of varying densities (1–220 k), a genomic identity-by-descent model (IBD-GS), using linkage analysis of sparse genome-wide markers, as well as a classical pedigree-based model. Reliabilities of the models were compared through 5-fold cross-validation. The traits studied were salmon lice (Lepeophtheirus salmonis) resistance (LR), measured as (log) density on the skin and fillet color (FC), with respective estimated heritabilities of 0.14 and 0.43. All genomic models outperformed the classical pedigree-based model, for both traits and at all marker densities. However, the relative improvement differed considerably between traits, models and marker densities. For the highly heritable FC, the IBD-GS had similar reliability as GBLUP at high marker densities (>22 k). In contrast, for the lowly heritable LR, IBD-GS was clearly inferior to GBLUP, irrespective of marker density. Hence, GBLUP was robust to marker density for the lowly heritable LR, but sensitive to marker density for the highly heritable FC. We hypothesize that this phenomenon may be explained by historical admixture of different founder populations, expected to reduce short-range lice density (LD) and induce long-range LD. The relative importance of LD/relationship information is expected to decrease/increase with increasing heritability of the trait. Still, using the ordinary GBLUP, the typical long-range LD of an admixed population may be effectively captured by sparse markers, while efficient utilization of relationship information may require denser markers (e.g., 22 k or more).