OATAO - Open Archive Toulouse Archive Ouverte Open Access Week

Genomic selection accounting for non-additive genetic effects in pig and corn crossbreeding schemes

Gonzalez Dieguez, David. Genomic selection accounting for non-additive genetic effects in pig and corn crossbreeding schemes. PhD, Pathologie, Toxicologie, Génétique et Nutrition, Institut National Polytechnique de Toulouse, 2020

(Document in English)

PDF (Author's version) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader


This thesis explores and develops methodology to exploit dominance or/and epistasis genetic effects on genomic selection models in pig and maize crossbreeding schemes. The Chapter 2 consisted of estimating and exploiting within-breed dominance variance through mate allocation strategies to maximize the overall genetic merit of the traits age at 100 Kg (AGE), backfat depth (BD) and average piglet weight per litter (APWL), in a French Landrace pig population. Maximizing total genetic values instead of breeding values in matings gave to the progeny an average advantage of 0.79 days, 0.04 mm, and 11.3 g for AGE, BD and APWL, respectively, but slightly reduced the expected additive genetic gain (e.g. 1.8 % for AGE). These results indicate that genomic mate allocation can improve the performance of the offspring without dramatically compromising the additive genetic gain. In Chapter 3, the effectiveness of mate allocation strategies and genomic evaluations, accounting for additive and dominance effects, to improve crossbred (CB) performance were investigated by simulation in a two-way pig crossbreeding scheme. Effects of the sources of information used in the genetic evaluation (only purebred (PB) data or PB and CB data), of several narrow and broad-sense heritability values, and of several options for mate allocation to produce the CB (mating at random, minimizing expected future inbreeding, or maximizing the expected total genetic value of crossbred animals) were evaluated. Selecting PB animals for PB performance yielded a genetic gain of 0.2 genetic standard deviations of the trait “CB performance” per generation, whereas selecting PB animals for CB performance doubled the genetic response. Mate allocation strategy resulted in a slight increase of the CB performance. When the genetic correlation between PB and CB is low, selecting PB animals for CB performance using CB information is a more efficient strategy to exploit heterosis and increase performance at the CB commercial level. In Chapter 4, the theory of hybrid genetic evaluation models from single-cross of pure lines (as in maize) was revisited in a genomic context. Covariance between hybrids due to additive substitution effects and dominance and epistatic deviations were analytically derived. Using SNP genotypes, it is possible to split specific combining ability (SCA) into dominance and across-groups epistasis, and to split general combining ability (GCA) into within-line additive effects and within-line additive by additive epistasis. A publicly available maize data set of Dent × Flint hybrids was analyzed. The proposed model was compared to other genomic models in terms of variance components estimation and predictive ability, including a model assuming a common effect of genes across origins. The study confirms that most variation in hybrids is accounted for by GCA, and that variances due to dominance and epistasis are small and have similar magnitudes. Models based on defining effects either differently (as it is traditionally done in maize) or identically across origins (as it is done in single breeds in livestock) resulted in similar predictive abilities for hybrids.

Item Type:PhD Thesis
Uncontrolled Keywords:
Institution:Université de Toulouse > Institut National Polytechnique de Toulouse - Toulouse INP (FRANCE)
Laboratory name:
Research Director:
Vitezica, Zulma and Tusell, Llibertat Palomero
Deposited On:28 Apr 2021 07:55

Repository Staff Only: item control page