Modeling heterogeneous (co)variances from adjacent-SNP groups improves genomic prediction for milk protein composition traits

Table 1 Heritability estimates and genome-wide correlations and covariances with total milk protein yield

Trait^a	h²	SE	Covariance	SE	Correlation	SE
\({\upalpha}\) _S1-CN	0.14	0.07	0.01	0.05	0.04	0.16
\({\upalpha}\) _S1-CN-8P	0.14	0.09	− 0.02	0.05	− 0.07	0.16
\({\upalpha}\) _S2-CN	0.33	0.09	− 0.08	0.06	− 0.16	0.12
\({{\upkappa}}\)-CN	0.69	0.09	0.06	0.05	0.09	0.07
G-\({{\upkappa}}\)-CN	0.41	0.09	0.0008	0.04	0.0006	0.10
\({\upalpha}\)-LA	0.15	0.09	0.05	0.05	0.15	0.16
\({{\upbeta}}\)-LG	0.52	0.10	0.04	0.05	0.07	0.09
Protein %	0.54	0.09	− 0.08	0.06	− 0.14	0.10

Heritability (h²) estimates were from the univariate GBLUP analysis; covariances and correlations are from the bivariate GBLUP model
^aProtein composition expressed as a fraction of the total milk protein percentage by weight wt (wt/wt), protein % expressed as percentage of the total milk yield; individual proteins comprise only the peaks identified as intact proteins and isoforms, i.e., \({\upalpha}\) _S1-CN (comprises \({\upalpha}\) _S1-CN 8P + 9P), \({\upalpha}\) _S2-CN (comprises \({\upalpha}\) _S2-CN 11P + 12P), \({{\upkappa}}\)-CN (comprises \({{\upkappa}}\)-CN G 1P + unglycosylated \({{\upkappa}}\)-CN 1P), where P = phosphorylated serine group. G-\({{\upkappa}}\)-CN = glycosylated-\({{\upkappa}}\)-CN; \({\upalpha}\) _S1-CN-8P = \({\upalpha}\) _S1-CN with 8 phosphorylated serine groups

ISSN: 1297-9686