Elahe Sanjari Banestani - Department of Animal Science, Faculty of Agriculture, Shahid Bahonar University of Kerman, Kerman, Iran
Ali Esmailizadeh - Department of Animal Science, Faculty of Agriculture, Shahid Bahonar University of Kerman, Kerman, Iran
Mehdi Moemen - Department of Surgical Sciences, School of Veterinary Medicine, University of Wisconsin-Madison, Madison, WI ۵۳۷۰۶, USA
Ahmad Ayatollahi Mehrgardi - Department of Animal Science, Faculty of Agriculture, Shahid Bahonar University of Kerman, Kerman, Iran
Morteza Mokhtari - Department of Animal Science, Faculty of Agriculture, University of Jiroft, P.O. Box ۳۶۴, Jiroft, Iran

The present study was performed to quantify a latent variable for body size (BS) from the five linear body measurements, including body length (BL), body height (BH), chest width (CW), chest girth (CG), and tube girth (TG). The study population consisted of N= ۵۵۷۳ Yorkshire pigs, ۵۹۲ individuals out of them were genotyped using a PorcineSNP۸۰ BeadChip. The body size latent variable was determined using Confirmatory Factor Analysis (CFA). Then, the accuracy of breeding values was obtained using pedigree-based best linear unbiased prediction (PBLUP), genomic best linear unbiased prediction (GBLUP), and single-step genomic best linear unbiased prediction (ssGBLUP) models. The overall fit indices, including standardized root mean square residual (SRMR), root mean square error of approximation (RMSEA), Tucker-Lewis Index (TLI), and comparative fit index (CFI) were obtained for the BS as ۰.۰۳, ۰.۰۹, ۰.۹۳, and ۰.۹۶, respectively which imply the adequacy of the considered model for BS construct. The performance of models was measured in a ۵-fold cross-validation with ۱۰ repeats to get a more accurate measure of the model's performance. The accuracy of models was compared via the correlation between predicted breeding values (PBV) and estimated breeding values (EBV) metric which was ۰.۳۷, ۰.۳۰, and ۰.۲۸ for PBLUP, ssGBLUP, and GBLUP, respectively. Furthermore, the goodness of fit is measured by the mean square of error (MSE) and Pearson's correlations r(y, ) between observed and predicted phenotypes. The lowest MSE and the highest Pearson's correlations were obtained under PBLUP while the highest MSE and the lowest Pearson's correlations were obtained under GBLUP. The obtained results showed the GBLUP method generally provided lower prediction accuracies than PBLUP and ssGBLUP methods, and also ssGBLUP generated lower prediction accuracy than traditional PBLUP. The performance of ssGBLUP and GBLUP was lower than expected mainly due to the small number of genotyped animals.

body dimension, Cross-Validation, genetic evaluation, latent variable, PIG