Skip to main content

Table 3 Accuracy of the best prediction models with haplotype additive values compared to the best SNP models

From: Haplotype genomic prediction of phenotypic values based on chromosome distance and gene boundaries using low-coverage sequencing in Duroc pigs

  Trait
AGW ADG BJS FCR LMA LMD BF TN
SNP accuracy for predicting phenotypic values \({\hat{\text{R}}}_{{0}_{\text{jp}}}=\text{corr}({\hat{\text{g}}}_{{0}_{\text{j}}},{\text{y}}_{0})\)
 A-only, Model-6 0.251 0.258 0.258 0.197 0.402 0.363 0.387 0.401
 A + D, Model-5 0.252 0.259 0.244 0.202 0.401 0.364 0.387 0.405
 A + D over A (%) 0.381 0.328 − 5.588 2.624 − 0.123 0.496 0.174 0.856
Best SNP prediction model (the SNP model in italic font, A-only or A + D)
 \({ \hat{\text{R}}}_{{0}_{\text{jp}}}=\text{corr}({\hat{\text{g}}}_{{0}_{\text{j}}},{\text{y}}_{0})\) 0.252 ± 0.032 0.259 ± 0.028 0.258 ± 0.033 0.202 ± 0.052 0.402 ± 0.042 0.364 ± 0.055 0.387 ± 0.061 0.405 ± 0.055
 \({\text{ R}}_{{0}_{\text{jp}}}={\text{R}}_{{0}_{\text{j}}}\sqrt{{\text{h}}_{\text{j}}^{2}}\) 0.285 ± 0.010 0.290 ± 0.009 0.156 ± 0.013 0.241 ± 0.012 0.432 ± 0.008 0.418 ± 0.010 0.407 ± 0.015 0.405 ± 0.010
 \({\text{ R}}_{{0}_{\text{j}}}=\text{corr}({\hat{\text{g}}}_{{0}_{\text{j}}},{\text{g}}_{{0}_{\text{j}}})\) 0.613 ± 0.010 0.624 ± 0.01 0.559 ± 0.016 0.579 ± 0.015 0.757 ± 0.004 0.722 ± 0.009 0.718 ± 0.010 0.707 ± 0.007
Haplotype prediction accuracy
 Best model H H H D + H H A + D + H A + D + H A + D + H
 Best blocking 500 kb 500 kb 100 kb 1 Mb Genes Genes 1 Mb Genes
 \({ \hat{\text{R}}}_{{0}_{\text{jp}}}=\text{corr}({\hat{\text{g}}}_{{0}_{\text{j}}},{\text{y}}_{0})\) 0.270 ± 0.029 0.276 ± 0.027 0.277 ± 0.026 0.212 ± 0.066 0.413 ± 0.043 0.371 ± 0.055 0.392 ± 0.062 0.406 ± 0.053
 Accuracy increase (%) 7.14 6.56 7.36 4.95 2.74 1.92 1.29 0.25
 \({\text{ R}}_{{0}_{\text{jp}}}={\text{R}}_{{0}_{\text{j}}}\sqrt{{\text{h}}_{\text{j}}^{2}}\) 0.292 ± 0.006 0.298 ± 0.005 0.178 ± 0.011 0.248 ± 0.014 0.431 ± 0.006 0.417 ± 0.010 0.413 ± 0.015 0.401 ± 0.010
 \({\text{ R}}_{{0}_{\text{j}}}=\text{corr}({\hat{\text{g}}}_{{0}_{\text{j}}},{\text{g}}_{{0}_{\text{j}}})\) 0.647 ± 0.005 0.650 ± 0.004 0.572 ± 0.012 0.549 ± 0.012 0.743 ± 0.004 0.710 ± 0.008 0.693 ± 0.009 0.695 ± 0.007
  1. \({\hat{\text{R}}}_{{0}_{\text{jp}}}\), observed accuracy of predicting phenotypic values; \({\text{R}}_{{0}_{\text{jp}}}\), theoretical accuracy of predicting phenotypic values; \({\text{R}}_{{0}_{\text{j}}}\), theoretical accuracy of predicting genotypic values; accuracy increase is the percentage increase in observed accuracy of predicting phenotypic values under the best haplotype model relative to the observed accuracy of the best SNP model (in italic font); A, SNP additive values; D, SNP dominance values; H, haplotype additive values; AGW, age at 100 kg live weight; ADG, daily gain; BJS, body judging score; FCR, feed conversion ratio; LMA, loin muscle area; LMD, loin muscle depth; BF, back fat thickness; TN, teat number