The impact of training on data from genetically-related lines on the accuracy of genomic predictions for feed efficiency traits in pigs

Abstract

Background Most genomic predictions use a unique population that is split into a training and a validation set. However, genomic prediction using genetically heterogeneous training sets could provide more flexibility when constructing the training sets in small populations. The aim of our study was to investigate the potential of genomic prediction of feed efficiency related traits using training sets that combine animals from two different, but genetically-related lines. We compared realized prediction accuracy and prediction bias for different training set compositions for five production traits. Results Genomic breeding values (GEBV) were predicted using the single-step genomic best linear unbiased prediction method in six scenarios applied iteratively to two genetically-related lines (i.e. 12 scenarios). The objective for all scenarios was to predict GEBV of pigs in the last three generations (~ 400 pigs, G7 to G9) of a given line. For each line, a control scenario was set up with a training set that included only animals from that line (target line). For all traits, adding more animals from the other line to the training set did not increase prediction accuracy compared to the control scenario. A small decrease in prediction accuracies was found for average daily gain, backfat thickness, and daily feed intake as the number of animals from the target line decreased in the training set. Including more animals from the other line did not decrease prediction accuracy for feed conversion ratio and residual feed intake, which were both highly affected by selection within lines. However, prediction biases were systematic for these cases and might be reduced with bivariate analyses. Conclusions Our results show that genomic prediction using a training set that includes animals from genetically-related lines can be as accurate as genomic prediction using a training set from the target population. With combined reference sets, accuracy increased for traits that were highly affected by selection. Our results provide insights into the design of reference populations, especially to initiate genomic selection in small-sized lines, for which the number of historical samples is small and that are developed simultaneously. This applies especially to poultry and pig breeding and to other crossbreeding schemes.

Genetic relationships between feed efficiency and gut microbiome in pig lines selected for residual feed intake

“…(2017) and Aliakbari et al. (2020). Briefly, the G0 individuals were obtained from artificial insemination of 30 sows with 30 boars in generation F0.…”

“…To increase the statistical power, given the high r g estimated in preliminary analyses between the traits measured in candidate and response animals, the phenotypic records were combined for both cohorts, after standardization of the records from candidates to selection to the variance of the corresponding trait of the response animals, as described in Aliakbari et al. (2020). Descriptive information of the five traits from G0 to G10 are given in Table 1.…”

Genomic Prediction of Average Daily Gain, Back-Fat Thickness, and Loin Muscle Depth Using Different Genomic Tools in Canadian Swine Populations

“…Additionally, the size of validation group in Duroc (774) was higher than Yorkshire (471) and Landrace (392–393), which could be an important factor in calculating the genomic prediction accuracy. The validation group size and selection of response variable are important factors on prediction accuracy, which have not been highlighted in the previous genomic prediction studies in pig ( Badke et al, 2014 ; Song et al, 2017 , 2019a ; Thekkoot et al, 2018 ; Zhang et al, 2018 ; Lopez et al, 2019 ; Aliakbari et al, 2020 ).…”

“…Several genomic prediction models of GBLUP ( Badke et al, 2014 ; Song et al, 2017 , 2019b , 2020 ; Jafarikia et al, 2018 ; Zhang et al, 2018 ), ssGBLUP ( Song et al, 2017 , 2019b ; Thekkoot et al, 2018 ; Hong et al, 2019 ; Lopez et al, 2019 ; Zhou et al, 2019 ; Aliakbari et al, 2020 ), and BayesC ( Esfandyari et al, 2016 ; Song et al, 2020 ) have frequently been used for prediction of various traits in the previous studies in swine. The main assumption of GBLUP method is based on the infinitesimal model (i.e., the genetic variation of the trait was explained by a large number of loci) ( Karaman et al, 2016 , 2018 ) that has been widely used in genomic evaluation, principally due to its ease of implementation, in developed countries’ breeding programs, such as Holstein cattle breeding programs in Canada 1 .…”

Identification of genomic regions affecting production traits in pigs divergently selected for feed efficiency

“…GWAS analyses were performed using the GEMMA software (version 0.97) [ 21 ] on response animals with their own phenotypes and their average genotypes from the parents. Phenotypes were adjusted for significant fixed effects and covariates (pen size, herd, sex, and contemporary groups for in vivo measurements, slaughter date as fixed effects, and slaughter age as covariate for traits recorded at the abattoir, and slaughter BW as covariate for carcBFT) using linear models as proposed in Aliakbari et al [ 22 ]. The resulting residues were integrated as phenotypes in GEMMA.…”