Whole-genome sequence data for selective breeding against infectious diseases in aquaculture

The aim of this project was to use genomic information to increase the genetic resistance to common diseases in Atlantic cod farming. We have obtained whole-genome sequence information from 111 individuals originating from the National cod breeding program, to build a data set with genetic (SNP) markers. We have identified 19 million markers in Atlantic cod and 1-3 million of these were used in the analysis of this project. Access to an improved version of all chromosomes of the reference genome has been essential for the analysis for this project. We have mapped genes for resistance against nodavirus (markers explain 35% of the genetic variation), which is a disease that is very common to many marine aquaculture species. Gene mapping for resistance to vibriosis, a common bacterial disease, points to a polygenic inheritance. Accuracy of selection has been compared for three information sources: pedigree, low density SNP chip and high density sequence data. Results show that sequence data had 20 and 4% higher accuracy of selection than pedigree and SNP chip information. We show an increasing discrepancy between traditional and genomic rates of inbreeding over time when rates of inbreeding are controlled on traditional pedigree information only and selection is on genomic breeding values. This shows that it is important to control rates of inbreeding at the genomic level for genomic selection. Historical effective population sizes were estimated to be 6 x 10-4 for both the North East Atlantic Cod and Norwegian Coastal Cod populations, and there was a small and temporary reduction 1 x 10-5 years ago. In addition, our data on gene frequencies from coastal and north east Atlantic cod generally show that they are mixed populations, and that some few very well defined regions of the genome have strong natural selection pressure.

Norway has taken a leading role in the sequencing of the Atlantic cod and Atlantic salmon genomes. The information content is enormous, but will only benefit us if we manage to exploit these data. This project will develop methodology to use genome sequen ces for selective breeding against infectious diseases in aquaculture. To this aim, we will sequence pooled DNA of cod challenge tested for vibriosis and viral nervous necrosis from the National Breeding Program, align this to the latest cod reference gen ome and estimate the effects of the DNA polymorphisms. Genetic values of parents will be calculated using these effects, and compared to conventional and sparse marker map based estimates. We will also map genes for resistance to vibriosis and viral nervo us necrosis, and use the sequence data to estimate historical and current effective population sizes. The effects of genome sequence data on aquaculture breeding are expected to be far-reaching, where individual breeding values of selection candidates may be calculated even in absence of own phenotype and pedigree recording. Hence, completely different breeding schemes may be envisaged due to sequence information, and such alternative schemes will be developed and tested within the project, together with methods that manage the inbreeding at the DNA level. In conclusion, the expected benefits from whole genome sequence data are large, and the project covers all aspects of this new data with respect to selective breeding in aquaculture.