Skip to main content
. 2021 Feb 12;12(2):266. doi: 10.3390/genes12020266

Figure 4.

Figure 4

Prediction accuracy and bias of GEBV/estimated breeding value (EBV) obtained using pedigree-based best linear unbiased prediction (PBLUP), single-step genomic best linear unbiased prediction (ssGBLUP), and the best single and window weighting ssGBLUP (WssGBLUP) for the studied traits. BFT, backfat thickness; CW, carcass weight; EMA, eye muscle area; MS, marbling score; YW, yearling weight. The default method for CW and YW at the second iteration, the constant method for MS and EMA at iteration 2 and 5, respectively, and nonlinear A with CT value of 1.25 and limiting the maximum SNP variance of 5 (nonlinearA_1.25_5) at iteration 2 for BFT were the best single weighting WssGBLUP method. The summed window (window size = 75) for CW and YW at iteration 3, The largest window (window size = 75) for BFT and MS at iteration 2, and the largest window (window size = 100) at the third iteration for EMA were the best window weighting WssGBLUP method.