Genomic selection is a modified form of marker-assisted selection in which the markers from the whole genome are used to estimate the genomic-estimated breeding value (GEBV). Several estimators are available to estimate GEBV. These estimators are able to capture either additive genetic effects or nonadditive genetic effects. However, there is hardly any procedure available that could capture both the effects simultaneously. Therefore, this study has been conducted to develop an integrated framework that is able to capture both additive and nonadditive effects efficiently. This integrated framework has been developed after evaluating existing additive and nonadditive models for marker selection. Furthermore, two efficient additive and nonadditive methods, that is, sparse additive models (SpAM) and Hilbert–Schmidt independence criterion least absolute shrinkage and selection operator (HSIC LASSO), have been combined to select both additive and nonadditive genetic markers for estimation of GEBV. The performance of the proposed framework has been evaluated on the basis of prediction accuracy, fraction of correctly selected features, and redundancy rate, along with standard error of mean for estimation of GEBV, compared with the individual performances of SpAM and HSIC LASSO separately. The newly developed framework is found to be satisfactory in terms of its performance and found to be robust for estimation of GEBV.

中文翻译：

---

选择用于基因组选择的附加和非附加遗传标记的综合框架。

基因组选择是标记辅助选择的一种改进形式，其中来自全基因组的标记用于估计基因组估计育种值 (GEBV)。有几个估算器可用于估算 GEBV。这些估计量能够捕捉加性遗传效应或非加性遗传效应。然而，几乎没有任何程序可以同时捕捉这两种效果。因此，本研究旨在开发一个能够有效捕捉加性和非加性效应的综合框架。在评估用于标记选择的现有加法和非加法模型后，开发了该集成框架。此外，两种有效的加法和非加法方法，即，稀疏加性模型 (SpAM) 和 Hilbert-Schmidt 独立准则最小绝对收缩和选择算子 (HSIC LASSO) 已结合以选择加性和非加性遗传标记来估计 GEBV。已根据预测精度、正确选择特征的比例和冗余率以及估计 GEBV 的均值标准误差评估了所提出框架的性能，并分别与 SpAM 和 HSIC LASSO 的单独性能进行了比较。发现新开发的框架在其性能方面令人满意，并且对于估计 GEBV 是稳健的。已根据预测精度、正确选择特征的比例和冗余率以及估计 GEBV 的均值标准误差评估了所提出框架的性能，并分别与 SpAM 和 HSIC LASSO 的单独性能进行了比较。发现新开发的框架在其性能方面令人满意，并且对于估计 GEBV 是稳健的。已根据预测精度、正确选择特征的比例和冗余率以及估计 GEBV 的均值标准误差评估了所提出框架的性能，并分别与 SpAM 和 HSIC LASSO 的单独性能进行了比较。发现新开发的框架在其性能方面令人满意，并且对于估计 GEBV 是稳健的。