Multivariate genome-wide association identify loci for complex growth traits by considering additive and over-dominance effects in hybrid Eucalyptus

• Genome-wide association studies are a powerful and widely used approach to decipher the genetic control of quantitative traits. One of the major challenges for traits in hybrid forest trees, such as hybrid Eucalyptus, is dissecting also non-additive effects for complex traits using a traditional linear mixed model. These non-additive effects, especially over-dominance effects, are one of most important hypotheses for the genetic basis of heterosis. In this study, we used a population including 949 F1 hybrids and their 174 parents, that were phenotyped for circumference at breast height and height at age of three years and six years, and also genotyped at 37,832 informative SNPs. Here we use and compare single-trait and multi-trait association models by accounting for additive and over-dominance effects, to evaluate genomic regions associated with the growth traits. For additive effect-based association model, nine significant SNPs were observed in multi-trait analyses, whereas only two unique SNPs were detected in single-trait analyses. These two SNPs were also identified in the multi-trait model. When evaluating over-dominance effects, 17 and 13 SNPs were identified from multi-trait and single-trait models, respectively. Moreover, more phenotypic variation can be explained by SNPs identified from multi-trait GWAS when including over-dominance effects. Overall, this study shows the added values of including over-dominance and considering multiple traits for identifying genomic regions that control traits of interest and that could contribute to heterosis in hybrids.

Subject headings

NATURVETENSKAP  -- Biologi -- Genetik (hsv//swe)

NATURAL SCIENCES  -- Biological Sciences -- Genetics (hsv//eng)

LANTBRUKSVETENSKAPER  -- Lantbruksvetenskap, skogsbruk och fiske -- Skogsvetenskap (hsv//swe)

AGRICULTURAL SCIENCES  -- Agriculture, Forestry and Fisheries -- Forest Science (hsv//eng)

Keyword

Eucalyptus grandis

E. urophylla

heterosis

multivariate mixed linear model

univariate mixed linear model

Publication and Content Type

vet (subject category)

ovr (subject category)