Contribution of rare and low-frequency whole-genome sequence variants to complex traits variation in dairy cattle

Table 3 Proportion of DRP variance explained using different methods

Traits	GREML-MS	REML-GRM	REML-PED	REML-PEDGRM
YIELD	0.860	0.845	0.923	0.941
MILK	0.872	0.844	0.887	0.927
PROT	0.858	0.847	0.943	0.963
FAT	0.854	0.840	0.898	0.914
MILKORG	0.679	0.703	0.811	0.816
MILKSP	0.719	0.715	0.748	0.840
LONG	0.630	0.606	0.884	0.881
MASTI	0.669	0.684	0.704	0.769
HEALTH	0.514	0.502	0.893	0.892
LEG	0.525	0.525	0.709	0.669
CALV	0.507	0.504	0.698	0.689
BIRTH	0.602	0.612	0.698	0.695
FERT	0.600	0.594	0.851	0.769
BODY	0.568	0.560	0.633	0.594
GROWTH	0.814	0.800	0.916	0.943
TEMP	0.406	0.403	0.645	0.645
NTM	0.847	0.839	–^a	–

GREML-MS refers to estimation using the GREML-MS method with imputed sequence variants partitioned into MAF classes. REML-GRM refers to estimation using 50 k SNPs with the REML-GRM model implemented in GCTA. REML-PED refers to using pedigree relationship in the REML-PED model implemented in DMU. REML-PEDGRM refers to fitting both 50 k SNPs and pedigree relationship in the REML-PEDGRM model implemented in DMU
^aThe model did not converge

ISSN: 1297-9686