The predictive ability of indirect genetic models is reduced when culled animals are omitted from the data

Table 4 Predictive abilities^a of the classical genetic model (CGM) and four indirect genetic models without (IGM) or with weighting of indirect genetic effects (\(IGM_{twg}\), \(IGM_{sbg}\), and \(IGM_{twsbg}\)) for six editing procedure scenarios^b

Scenario	Model	\(cor\left( {Y_{c,i} ,a_{Di} } \right)\)	\(cor\left( {Y_{c,i} ,a_{Igrp} } \right)\)	\(cor\left( {Y_{c,i} ,a_{Di,Igrp} } \right)\)
full	\({\text{CGM}}\)	0.248^a	–	–
	\({\text{IGM}}\)	0.248^a	0.038^a	0.251^a
	\(IGM_{twg}\)	0.248^a	0.029^b	0.251^a
	\(IGM_{sbg}\)	0.248^a	0.032^c	0.251^a
	\(IGM_{twsbg}\)	0.248^a	0.015^d	0.250^b
noPhen	\({\text{CGM}}\)	0.248^a	–	–
noPhen	\({\text{IGM}}\)	0.248^a	− 0.074	0.248
testdays	\({\text{CGM}}\)	0.248^a	–	–
	\({\text{IGM}}\)	0.248^a	0.034^a	0.251^a
	\(IGM_{twg}\)	0.248^a	0.027^b	0.251^a
	\(IGM_{sbg}\)	0.248^a	0.026^b	0.251^a
	\(IGM_{twsbg}\)	0.248^a	0.013^c	0.249^b
noPhenGrp	\({\text{CGM}}\)	0.256^a	–
noPhenGrp	\({\text{IGM}}\)	0.256^a	0.026	0.258
noPhenGrp15	\({\text{CGM}}\)	0.250^a	–	–
	\({\text{IGM}}\)	0.250^a	0.028^a	0.252^a
	\(IGM_{twg}\)	0.250^a	0.021^b	0.252^a
	\(IGM_{sbg}\)	0.250^a	0.017^c	0.251^b
	\(IGM_{twsbg}\)	0.250^a	0.007^d	0.250^c
predPhen	\({\text{CGM}}\)	0.249^a	–	–
	\({\text{IGM}}\)	0.248^b	0.038^a	0.251^a
	\(IGM_{twg}\)	0.249^a	0.008^b	0.250^b
	\(IGM_{sbg}\)	0.250^c	0.007^b	0.250^b
	\(IGM_{twsbg}\)	0.249^a	− 0.005^c	0.248^c

Different letters indicate significant (P ≤ 1%) differences between models within the columns and scenarios
^aPredictive abilities were calculated as the correlation between predicted breeding values and corrected phenotypes for: individual direct breeding values, \(a_{Di}\), the sum of the indirect breeding values of the group mates, \(a_{Igrp}\), and the combination of direct, individual breeding value and the sum of the indirect breeding values of the group mates,\(a_{Di,Igrp}\)
^bThe full scenario with no editing based on culled pigs or pigs without phenotypes and five alternative scenarios: noPhen (individually culled pigs or pigs without phenotypes were omitted), testdays (pigs culled within the first 28% of test days were omitted), noPhenGrp (complete groups with culled pigs or pigs with missing phenotypes omitted), noPhenGrp15 (groups with more than 15% culled pigs or missing phenotypes were omitted), and predPhen (predicted breeding values assigned to pigs without phenotypes including culled pigs)

ISSN: 1297-9686