Table 4.
Agreement on the top 10-ranked broilers selected across the different genome-enabled prediction approaches and for the various sub-sampling of the entire training dataset
| Predictive Approach | Training dataset size (%) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 3 | 5 | 7 | 10 | 15 | 20 | 30 | 40 | 50 | 60 | 70 | 80 | 90 | 100 | |
| BRR x Bayes Cπ | 91.5 | 94.5 | 95.0 | 94.4 | 95.9 | 93.9 | 93.2 | 90.8 | 90.3 | 87.6 | 86.6 | 89.1 | 87.6 | 87.9 | 88.6 |
| BRR x DNN | 57.8 | 66.9 | 64.9 | 79.3 | 78.4 | 82.6 | 55.5 | 79.8 | 76.6 | 73.1 | 72.9 | 77.1 | 78.5 | 75.9 | 76.5 |
| BRR x BRR-WT | 28.2 | 37.3 | 44.5 | 50.1 | 57.7 | 62.8 | 68.5 | 74.3 | 78.8 | 81.5 | 83.3 | 84.6 | 86.4 | 86.1 | 87.9 |
| BRR x Bayes Cπ-WT | 28.3 | 36.8 | 44.9 | 50.5 | 58.1 | 63.4 | 64.7 | 72.1 | 75.8 | 76.4 | 79.5 | 80.1 | 80.6 | 83.0 | 83.5 |
| Bayes Cπ x DNN | 57.6 | 65.7 | 64.2 | 78.8 | 77.9 | 80.9 | 55.3 | 76.9 | 74.8 | 69.4 | 71.6 | 74.9 | 75.2 | 72.2 | 74.2 |
| Bayes Cπ x BRR-WT | 28.5 | 37.3 | 44.0 | 50.7 | 56.4 | 62.5 | 67.6 | 73.3 | 76.8 | 79.1 | 79.8 | 82.2 | 82.0 | 81.8 | 82.6 |
| Bayes Cπ x Bayes Cπ-WT | 28.7 | 36.9 | 44.2 | 50.9 | 57.3 | 63.4 | 64.1 | 73.2 | 75.8 | 78.7 | 80.4 | 81.3 | 81.3 | 83.3 | 85.2 |
| DNN x BRR-WT | 28.2 | 37.5 | 38.0 | 47.3 | 55.3 | 58.8 | 47.4 | 66.5 | 68.5 | 68.3 | 69.7 | 72.8 | 73.6 | 72.2 | 72.3 |
| DNN x Bayes Cπ-WT | 28.1 | 36.5 | 38.1 | 47.4 | 55.4 | 58.8 | 46.3 | 64.4 | 67.8 | 65.2 | 67.7 | 71.0 | 71.0 | 71.2 | 70.1 |
| BRR-WT x Bayes Cπ-WT | 94.1 | 96.2 | 93.3 | 93.9 | 94.5 | 93.9 | 89.5 | 88.9 | 88.7 | 85.7 | 89.6 | 88.6 | 88.1 | 88.3 | 88.2 |