| Summary: | Nowadays, molecular information has changed the way genetic improvement is per-formed, increasing the response to selection from the traditional approach, based only on genealogical information using best linear unbiased prediction (BLUP). In this study, breeding values accuracy obtained using molecular information (using single nucleotide polymorphism (SNP) arrays chip at different densities and whole genome sequence) ware compared by cross validation with BLUP. Three characters of economic interest to the salmon industry were used: body weight at harvest (BW) and resistance to two diseases (caused by pathogens Piscirickettsia salmonis (SRS) and Caligus rogercresseyi (CAL)). The effect of genetic architecture (heritability 'h2'), different trait types (continuous/discrete) and family structure were also taken into account. A typical scheme of fish farming based on families of siblings was simulated: 10 and 50 families with 500 and 100 full-sibs, respectively. Results showed a law of diminishing returns in genomic accuracy when increasing in SNPs chip density. Generally, the use of high number of SNPs or whole genome sequence may not be advantageous compared to medium density SNPs chip (besides the costs involved). The family structure had a clear effect on genomic accuracy for variability in the population. The best genomic evaluation accuracies were observed when few families (10) with few founder individuals (many females for one male) were considered. The relative improvements when using molecular information compared with the accuracy reached by BLUPs method was higher for BW character, continuous of high h2 (0,4), reaching an improvement in accuracy of 24%. For the CAL resistance character, continuous of low h2 (0,1), a maximum 9% of improvement in accuracy is obtained. The SRS resistance character, a binary character with h2 of 0,18, had a loss of accuracy in comparison with continuous character which (on the same level h2), have a 10% of improvement in accuracy. In the case of many families (50), ...
|
|---|