View Full Table | Close Full ViewTable 1.

Broad-sense heritability (H2) of agronomic traits evaluated in Cycle 29 Semi-Arid Wheat Screening Nursery.

 
Trait H2
Yield (irrigated) 0.62
Yield (drought) 0.62
Thousand kernel weight 0.95



View Full Table | Close Full ViewTable 2.

Prediction accuracy for agronomic traits in the Cycle 29 Semi-Arid Wheat Screening Nursery using different marker imputation methods. Cross-validation with seven folds was used and sister lines from the same cross were grouped in the same fold. No significant differences in prediction accuracy were observed among the imputation methods.

 
Trait MVN-EM† RF Mean Het
Yield (irrigated) 0.32 0.32 0.28 0.33
Yield (drought) 0.42 0.40 0.45 0.44
Thousand kernel weight 0.33 0.34 0.38 0.36
Days to heading 0.35 0.36 0.37 0.37
Imputation methods are multivariate normal expectation maximization (MVN-EM), random forest (RF), mean, and heterozygote (Het).



View Full Table | Close Full ViewTable 3.

Prediction accuracy and bias for simulated phenotypes.

 
Complete† MVN-EM Mean Het
Accuracy 0.477 a 0.466 b 0.457 c 0.444 d
Regression coefficient‡ 1.03 (0.02) 0.99 (0.02) 1.19 (0.02) 1.31 (0.03)
Imputation methods are complete genotypic data, multivariate normal expectation maximization (MVN-EM), mean, and heterozygote (Het). Significant differences between imputation methods are shown by letters.
Regression of true breeding value on genomic estimated breeding value with standard error shown in parenthesis.



View Full Table | Close Full ViewTable 4.

Cross validation prediction accuracy for genomic selection models using genotyping-by-sequencing (GBS) and Diversity Array Technology (DArT) markers on 254 elite breeding lines from the Cycle 29 Semi-Arid Wheat Screening Nursery, CIMMYT. Letters denotes significant differences within traits.

 
Marker type and number
Trait DArT 2000† GBS 2000‡ GBS 35000§
Yield (irrigated) 0.13 b 0.25 ab 0.32 a
Yield (drought) 0.18 b 0.35 a 0.42 a
Thousand kernel weight 0.28 a 0.26 a 0.33 a
Days to heading 0.20 b 0.34 a 0.33 a
Number of markers: 1729 for DArT, 1827 for GBS (2K), and 34,749 for GBS (35K).
Markers with up to 20% missing data.
§Markers with up to 80% missing data.