Estimation of Error Variance in Genomic Selection for Ultrahigh Dimensional Data

Estimation of error variance in the case genomic selection is a necessary step to measure accuracy model. For selection, whole-genome high-density marker data used where number markers always larger than sample size. This makes it difficult estimate because ordinary least square estimation technique cannot be datasets parameters greater individuals (i.e., p > n). In this article, two existing methods, viz. Refitted Cross Validation (RCV) and kfold-RCV, were suggested for such cases. Moreover, by considering limitations above new Bootstrap-RCV Ensemble method, have been proposed. Furthermore, an R package “varEst” has developed, which contains four different functions implement these methods Least Absolute Shrinkage Selection Operator (LASSO), Squares Regression (LSR) Sparse Additive Models (SpAM). The performances algorithms evaluated using simulated real datasets.