Using imputed whole-genome sequence data to improve the accuracy of genomic prediction for parasite resistance in Australian sheep

Table 3 Number of selected variants and selection criteria

Scenario	Selection criteria	Selected variants^a	Accuracy	Slope
Based only on GWAS
Scenario 1	\(50{\text{k}} + {\text{top}}_{{{\text{GWAS}}.{\text{seq}}}}\)	3913	0.21 ± 0.01	0.86 ± 0.01
Scenario 2	\(50{\text{k}} + {\text{top}}_{{{\text{GWAS}}.{\text{HD}}}}\)	226	0.20 ± 0.01	0.88 ± 0.01
Based only on RHM (1 Mbp)
Scenario 3	\(50{\text{k}} + {\text{top}}_{{{\text{RHM}}.{\text{seq}}}}\)	26,808	0.21 ± 0.01	0.90 ± 0.01
Based only on RHM (250 kbp)
Scenario 4	\(50{\text{k}} + {\text{top}}_{{{\text{RHM}}.{\text{seq}}}}\)	11,507	0.22 ± 0.01	0.91 ± 0.01
Scenario 5	\(50{\text{k}} + {\text{top}}_{{{\text{RHM}}.{\text{HD}}}}\)	992	0.22 ± 0.01	0.89 ± 0.01
Based on both RHM (250 kbp) and GWAS
Scenario 6	\(50{\text{k}} + {\text{top}}_{{{\text{GWAS}}.{\text{seq}}\left( {{\text{within}}\,{\text{RHM}}} \right)}}\)	413	0.25 ± 0.01	0.94 ± 0.01
Scenario 7	\(50{\text{k}} + {\text{top}}_{{{\text{GWAS}}.{\text{HD}}\left( {{\text{within}}\,\,{\text{RHM}}} \right)}}\)	49	0.23 ± 0.01	0.97 ± 0.01

In each scenario, a GRM from the 50k was fitted with a GRM from the selected variants
The selected variants were \({\text{top}}_{{{\text{GWAS}}.{\text{seq}}}}\): all variants that passed GWAS \(- log_{10} \left( p \right)\) threshold of 3, \({\text{top}}_{{{\text{GWAS}}.{\text{HD}}}}\): all HD variants that passed GWAS \(- log_{10} \left( p \right)\) threshold of 3, \({\text{top}}_{{{\text{RHM}}.{\text{seq}}}}\): all sequence variants within RHM windows that passed \(- log_{10} \left( p \right)\) threshold of 3, \({\text{top}}_{{{\text{RHM}}.{\text{HD}}}}\): all HD variants within RHM windows that passed \(- log_{10} \left( p \right)\) threshold of 3, \({\text{top}}_{{{\text{GWAS}}.{\text{seq}}\left( {{\text{within}}\,{\text{RHM}}} \right)}}\): only sequence variants that passed GWAS \(- log_{10} \left( p \right)\) threshold of 3 in RHM windows with \(- log_{10} \left( p \right) \ge 3\), \({\text{top}}_{{{\text{GWAS}}.{\text{HD}}\left( {{\text{within}}\,{\text{RHM}}} \right)}}\): only HD variants that passed GWAS \(- log_{10} \left( p \right)\) threshold of 3 in RHM windows with \(- log_{10} \left( p \right) \ge 3\)
^aNumber of selected variants after LD pruning

ISSN: 1297-9686