Skip to main content

Table 3 Predictive performance of GBLUP and MGBLUP for the LOLO scheme

From: Metabolomic-genomic prediction can improve prediction accuracy of breeding values for malting quality traits in barley

Trait

GBLUPg

MGBLUPg

cor

reg

cor

reg

FS

0.18

1.00

0.18

1.04

EY

0.32

1.02

0.32

1.07

WC

0.46

1.05

0.46

1.05

BG

0.38

1.09

0.38

1.15

WV

0.34

1.13

0.34

1.17

  1. GBLUP  genomic best linear unbiased prediction; MGBLUP  metabolomic-genomic best linear unbiased prediction; LOLO  leave one line out
  2. FS  filtering speed, EY  extract yield, WC  wort color, BG  beta-glucan, WV  wort viscosity
  3. GBLUPg is GBLUP incorporating genotypes on lines in VP, and MGBLUPg is similarly defined
  4. “cor” columns show the correlations between predicted breeding values and corrected phenotypes for lines in VP, and “reg” columns show the regression coefficients of corrected phenotypes on predicted breeding values. For all traits, none of the differences between the two correlations are statistically significant using a Hotelling-Williams t-test at a 5% level. Standard error on regression coefficient is 0.01 in all cases
Back to article page

Genetics Selection Evolution

ISSN: 1297-9686