Table 1.
Percentage of GBS concordance and call rates (in parentheses)
| Reference | Total | Homozygotes | Heterozygotes | |
|---|---|---|---|---|
| GBS1 Direct Read | 50 K | 98.88 (33.81) | 99.03 (33.72) | 45.09 (0.09) |
| 600 K | 98.99 (35.58) | 99.21 (35.47) | 28.67 (0.11) | |
| GBS2 Cornell Imputation | 50 K | 96.04 (91.56) | 98.66 (88.79) | 12.51 (2.78) |
| 600 K | 95.50 (93.41) | 98.69 (90.14) | 7.75 (3.28) | |
| GBS3 Beagle Imputation | 50 K | 93.04 (a91.56) | 93.23 (91.30) | 30.54 (0.26) |
| 600 K | 92.84 (a93.41) | 93.07 (93.12) | 22.50 (0.29) | |
| GBS4 Beagle Imputation on the missing data and heterozygotes after TASSEL Imputation (GBS2) | 50 K | 96.46 (a97.64) | 96.46 (97.63) | < 0.01 (< 0.01) |
| 600 K | 96.21 (a99,97) | 96.21 (> 99.99) | < 0.01 (< 0.01) | |
| GBS5 Compilation of Homoz. genotypes from TASSEL Imputation (GBS2) and Imputation by Beagle for Other Data (GBS3) | 50 K | 96.25 (a97.65) | 96.36 (97.47) | 39.07 (0.18) |
| 600 K | 95.98 (a99.97) | 96.11 (99.78) | 32.02 (0.22) |
The 50 K and 600 K SNP-arrays were considered as reference genotypes. a After Beagle inference of missing data, the call rate was 100%. Here the call rate is < 100% because the comparison was made against the 50 K and the 600 K that include few missing data. For GBS3, the remaining missing genotypes in GBS2 were also excluded to obtain comparable results