Table 5 Window size, log-likelihood, residual variance (\({\varvec{\upsigma}}_{{\mathbf{e}}}^{2}\)), additive variance (\({\varvec{\upsigma}}_{{\mathbf{a}}}^{2}\)) and intercept (\({\varvec{\upmu}}\)) with the best log-likelihood using real data and different methods
From: Study of the optimum haplotype length to build genomic relationship matrices
Trait | Method | DW | SW | TMS | \({\mathbf{G}}\) |
|---|---|---|---|---|---|
Scrotal circumference | Window size | 8 | 7 | 15 | – |
Log-likelihood | −1325.48 | −1325.60 | −1326.95 | −1330.15 | |
Additive variance | 5.29 | 5.44 | 5.02 | 3.20 | |
Residual variance | 2.03 | 2.11 | 2.32 | 2.44 | |
Age at puberty | Window size | 7 | 8 | 11 | – |
Log-likelihood | −4401.20 | −4402.31 | −4400.77 | −4405.25 | |
Additive variance | 11273.73 | 11052.93 | 11853.68 | 6816.86 | |
Residual variance | 4902.01 | 4694.35 | 5405.72 | 5729.85 | |
Weight at first corpus luteum | Window size | 7 | 2 | 3 | – |
Log-likelihood | −3493.23 | −3493.62 | −3492.76 | −3494.73 | |
Additive variance | 1451.34 | 2758.13 | 3184.40 | 900.60 | |
Residual variance | 750.88 | 804.50 | 845.43 | 833.38 |