Comparison of methods for deriving phenotypes from incomplete observation data with an application to age at puberty in dairy cattle

Table 3 Correlations between estimated breeding values (EBVs) from scenario one relative to those from scenarios two to eight for the three different methods used to derive age at first blood plasma progesterone elevation (AGEP4) phenotypes

Scenario		Correlations between EBVs relative to scenario one (EML)
Scenario		CAT	AGEVISIT	AUG
1	Early, mid and late visit (EML)	1.00	1.00	1.00
2	Early, mid or late visit (E/M/L)	0.67	0.65	0.75
3	Early visit only (E)	0.83	0.69	0.87
4	Mid visit only (M)	0.76	0.74	0.83
5	Late visit only (L)	0.74	0.75	0.83
6	Early and mid visit (EM)	0.94	0.95	0.95
7	Mid and late visit (ML)	0.90	0.95	0.94
8	Early and late visit (EL)	0.93	0.94	0.95

CAT: The number of the visit where the animal was first observed to have elevated BP4; AGEVISIT: Age in day at the visit the animal was first observed to have elevated BP4; and AUG: The continuous variable AGEP4 sampled from between the known lower and upper bounds. Correlations include within breed EBVs for sires that had > 5 daughters with an AGEP4 phenotype (n = 103). Scenarios are described in Fig. 1, and analysis methods are described in Table 2

ISSN: 2049-1891