Open AccessComparison of random regression models, traditional animal model and with the inclusion of molecular markers in the estimation of genetic parameters in Colombian Holstein cattle
Author(s) -
Daniel Cardona-Cifuentes,
Albeiro López Herrera,
Luis Gabriel González Herrera,
Mario Muñoz,
José Julián Echeverri-Zuluaga
Publication year - 2021
Publication title -
semina. ciências agrárias
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.268
H-Index - 25
eISSN - 1679-0359
pISSN - 1676-546X
DOI - 10.5433/1679-0359.2021v42n3p1303
Subject(s) - best linear unbiased prediction , heritability , selection (genetic algorithm) , statistics , biology , dairy cattle , random effects model , holstein cattle , animal model , estimation , microbiology and biotechnology , mathematics , genetics , computer science , machine learning , medicine , engineering , meta analysis , systems engineering , endocrinology
The use of molecular markers to identify desirable genes in animal production is known as marker-assisted selection. The traditional genetic evaluation model uses the BLUP methodology; when genetic markers are included in the evaluation model, the methodology is known as M-BLUP. In contrast, random regression models (RRM), unlike the models based on production at 305 days, consider factors that change for each animal from one test to another. The objective of this study was to compare variance components, genetic parameters and breeding values for milk production, protein percentage and somatic cell score in Colombian Holstein cattle using BLUP, M-BLUP and RRM. For the estimation of genetic parameters and values, 2003 lactations corresponding to 1417 cows in 55 herds were used, and effects of the order of delivery, herd, and contemporary group were included. The three traits presented greater heritability under the MBLUP model: 0.44 for protein percentage, 0.27 for milk production and 0.28 for somatic cell score. This was because the genetic variance was greater when M-BLUP was used, which allowed a greater accuracy of the breeding value estimation in the three traits. Therefore, the model that includes information on molecular markers is more suitable for genetic evaluation in Colombian Holstein cattle.