Skip to main content

Table 5 Prediction accuracies for wood quality and growth traits estimated with cross-validation sets built with full-sib families, using pedigree (CV2-A), ridge regression (CV2-RR) or both (CV2-RRA) sources of information with all SNPs ( m= 6,932) and various subpopulations of individuals

From: Genomic selection accuracies within and between environments and small breeding groups in white spruce

   Cross-validation
Trait Subpopulation* Pedigree (CV2-A) Markers (CV2-RR) Markers and pedigree (CV2-RRA) Relative efficiency markers/pedigree (%)
ADEN BG1 0.57 0.59 0.62 104
  BG1 → BG2 0.10 0.10 0.19
  S1 0.71 0.63 0.69 89
  S1 → S2 0.59 0.53 0.57 90
AMFA BG1 0.55 0.36 0.47 65
  BG1 → BG2 0.13 0.04 0.05
  S1 0.62 0.52 0.61 84
  S1 → S2 0.52 0.41 0.50 79
HT17 BG1 0.44 0.33 0.42 75
  BG1 → BG2 −0.01 −0.01 −0.01
  S1 0.66 0.55 0.64 83
  S1 → S2 0.33 0.22 0.29 67
DBH17 BG1 0.30 0.29 0.34 97
  BG1 → BG2 0.13 −0.06 −0.07
  S1 0.53 0.48 0.53 91
  S1 → S2 0.23 0.15 0.20 65
MSE   0.081 0.075 0.074
  1. ADEN: average wood density; AMFA: average microfibril angle; HT17: 17-year height; DBH17: 17-year diameter at breast height.
  2. *BG1: both training and testing sets from breeding group 1; BG1 → BG2: training sets from breeding group 1 and testing sets from breeding group 2; S1: both training and testing sets from site 1; S1 → S2: training sets from site 1 and testing sets from site 2.
  3. Mean standard error over cross-validation iterations and subpopulations.