Multivariate approach for young bull selection from a performance test using multiple traits of economic importance

Abstract

This study used multivariate statistics to identify clusters of animals with similar expected progeny difference (EPD) and also identify leading traits that discriminate between bulls. Various linear selection indices based on specific selection criteria were proposed. Records were collected from 880 young Nelore bulls submitted to performance testing in central Brazil between 2001 and 2012. Pre-weaning average daily gain and weights at 210 days with direct and maternal effects were used in the analysis, in addition to post-weaning weight, scrotal circumference at 365 and 450 days, carcass finish and rib eye area. EPDs were classified into three groups, and the EPD means of two of these groups stood out and were considered important based on principal component analysis that associated higher values of direct EPD of weights, average daily weight gain and scrotal circumference. The EPDs for weight at 210, 365 and 450 days, pre- and post-weaning daily gain and scrotal circumference at 365 and 450 days were major causes of variation. Each linear selection index proposed (SI₁, SI₂, SI₃, SI₄ and SI₅) defined a specific approach meaning that a different selection index should be used depending on breeding goals and selection criteria.

Principal component and cluster analyses of production and fertility traits in Red Sindhi dairy cattle breed in Brazil

“…and fertility (Karacaören and Kadarmideen 2008;Buzanskas et al 2013;Jolliffe and Cadima 2016;Lopes et al 2016). In addition, multivariate analyses are useful for addressing relevant decisions that reach male and female offspring with higher average performance related to previous generations by heterosis, increasing the variability of the population (Gianola and Sorensen 2004;Lopes et al 2013;Moraes et al 2015;Fraga et al 2016). Moreover, multivariate statistical techniques might reveal relationships that would not be possible with univariate statistical techniques (Usai et al 2006;Bolormaa et al 2010).…”

Integrating Casein Complex SNPs Additive, Dominance and Epistatic Effects on Genetic Parameters and Breeding Values Estimation for Murciano-Granadina Goat Milk Yield and Components

“…Out of the 48 SNPs studied, a total of 40 SNPs contributed to the seven-dimensional model in a meaningful way (factor loadings>|0.5| for CATPCA), then the different components (PC1, PC2, PC3, PC4, PC5, PC6, and PC7) were best described by the SNPs highlighted included in the red rectangle in Figure 1. SNP7, 9,11,21,27,30,33, and were discarded given they did not participate in any of the dimensions identified (confounding or variance explaining redundant SNPs).…”

“…Once superior individuals have been detected, identifying the genetic associations (dominance and additive effects) and the epistatic genetic relationships between traits of economic importance is essential as it may optimize the profitability of selection policies. In this context, multivariate analysis would allow us to address the decisions that may define the highest average-performing offspring in relation to the values of previous generations of heterosis and the increase in variability of the population [10,11].…”

Bivariate analysis for the improvement of genetic evaluations with incomplete records in Charolais cattle

“…El sesgo por la selección (sacrificio) de animales, puede ser en parte disminuida por modelos genéticos multivariados (4,5).…”