Relevance of Pedigree, Historical Data, Dominance, and Data Unbalance for Selection Efficiency

Abstract

Th e objective of this study was to assess the impact of pedigree, historical data, dominance, and data unbalance on the estimation and precision of genetic variances and breeding values and on the selection effi ciency in annual crop breeding. Expansion volume and grain yield from 12 trials of inbred progeny and four tests of non-inbred families were used in the analyses. Th e S 1 to S 5 progeny trials were designed as incomplete blocks, the S 6 progeny trials were designed as complete blocks, and the half-and full-sib family trials were designed as lattices. Th e half-sib, full-sib, and inbred family models were fi tted in across-generation analyses. One complete and four reduced models were used to assess the relevance of pedigree, historical data, and dominance. Simulated plot losses of 30% in the half-and full-sib progeny trials were used to study the infl uence of data unbalance. All analyses were performed using ASReml. Ignoring pedigree information or ancestor data and simulating plot losses determined relevant biases in estimating the additive and dominance variances, marked reduction in the precision of the predicted breeding values, signifi cant changes in the classifi cation of the breeding values, and errors in identifying superior individuals, i.e., a signifi cant reduction in the selection effi ciency. In contrast, excluding dominance had no signifi cant eff ect on either the ranking of breeding values or selection effi ciency. Our results revealed that best linear unbiased prediction including pedigree and historical data, based on a model with dominance, is the ideal method for genetic evaluation by plant breeders even when lost records are considered.

Multi-trait multi-environment Bayesian model reveals G x E interaction for nitrogen use efficiency components in tropical maize

“…We have chosen to adopt the MTME model for the inbred lines selection since a combined analysis takes more data into consideration and consequently is often more accurate to predict breeding values than single models [ 37 , 38 ]. According to [ 39 ], the increase in accuracy with the use of multi-trait best linear unbiased prediction (BLUP) analysis compared with single-trait analysis is proportional to the difference between the genetic and environmental correlations of the analyzed traits.…”

Parentesco na seleção para produtividade e teores de óleo e proteína em soja via modelos mistos

“…Portanto, quando se consideram condições experimentais limitantes, ou na seleção de multicaracterísticas, o emprego de modelos mistos com informação de parentesco pode ser o mais recomendado (Viana et al, 2010b(Viana et al, , 2012. Em termos práticos, a manutenção do registro de genealogias para progênies em teste pode demandar maior esforço, o que influencia a opção do uso destas informações na seleção.…”

“…Entretanto, no melhoramento de espécies anuais, seu uso não é tão frequente, embora haja um número crescente de relatos, em diferentes espécies, como milho (Viana et al, 2010a(Viana et al, , 2011(Viana et al, , 2012, feijão (Baldissera et al, 2012), arroz (Borges et al, 2012) e trigo .…”

Adaptability and stability analyses of plants using random regression models

“…When using trials with unbalanced data, or just a sample of the cultivars, the prediction accuracy estimates tend to be lower, and the model may not be efficient in evaluating the performance of the cultivars [ 72 , 73 ]. Viana et al [ 74 ] working with unbalanced data, showed that the lack of data resulted in relevant effects on the estimation of genetic variances, the accuracy of prediction and the ranking of predicted genetic values, which can affect the efficiency of selection.…”