Table 8

Summary of the results* of the test for equality of predictive ability across groups
Splitting Within-family Across-family
Trait W6W WGS BL %CD8+ CD4+/ CD8+ W6W WGS BL %CD8+ CD4+/ CD8+
emBayesB 0.42 0.30 0.26 0.28 0.05 0.22 0.47 0.33 0.06 0.03
RR_GBLUP 0.35 0.41 0.27 0.19 0.11 0.30 0.36 0.28 0.10 0.09
SS_BY 0.32 0.36 0.16 0.13 0.07 0.40 0.30 0.20 0.09 0.03
SS_ABS 0.39 0.34 0.28 0.04 0.17 0.20 0.28 0.34 0.06 0.06
RKHS 0.38 0.33 0.33 0.24 0.06 0.30 0.41 0.28 0.28 0.06
SVR 0.41 0.34 0.33 0.25 0.08 0.28 0.39 0.35 0.30 0.05
BayesCpi 0.37 0.28 0.27 0.28 0.07 0.31 0.38 0.25 0.12 0.07
BayesC 0.34 0.44 0.25 0.27 0.08 0.29 0.37 0.26 0.19 0.06
LASSO 0.45 0.29 0.24 0.24 0.10 0.30 0.36 0.38 0.10 0.03
RF 0.40 0.35 0.32 0.16 0.07 0.29 0.62 0.27 0.16 0.13

*P-values for the chi-square test for equality of predictive ability across groups (average of 10 replicates). In order to reduce the influence of discrepant replicates, the average of log10(p-value) was calculated for each trait and method, and then back-transformed to the original scale. Animals were clustered into 4 groups based on genetic distance between them.

Trait = weight at 6 weeks (W6W), weight growth slope (WGS), body length (BL), percentage of CD8+ cells (%CD8+), ratio between CD4+ and CD8+ cells (CD4+/CD8+). Splitting= splitting strategy in cross-validation (within or across-family).

Neves et al.

Neves et al. BMC Genetics 2012 13:100   doi:10.1186/1471-2156-13-100

Open Data