Table 6. The selected consensus artificial SNPs from the simulated data set.
| Artificial SNPs ID | SNP’s significance in the models | ||||
|---|---|---|---|---|---|
| RF with Stepwise | RF with Ridge regression | RF with LASSO | RF with Boruta (P = 0.01) | Consensus (3/4) | |
| X1306 | ✔(4.15e-05) | ✔(0.0003) | ✔(1.73e-05) | ✔ | ✔ |
| X92 | ✔(0.0004) | ✔(0.003) | ✔(0.0014) | × | ✔ |
| X1112 | ✔(0.0017) | ✔(0.00204) | ✔(0.0013) | ✔ | ✔ |
| X808 | ✔(0.0013) | × | ✔(0.0073) | × | |
| X859 | ✔(0.0034) | ✔(0.0061) | ✔(0.001) | × | ✔ |
| X263 | ✔(0.0021) | ✔(0.0151) | ✔(0.0061) | ✔ | ✔ |
| X829 | × | ✔(0.0065) | ✔(0.009) | × | |
| X1203 | × | ✔(0.0171) | × | ✔ | |
| X242 | × | ✔(0.0075) | ✔(0.003) | ✔ | ✔ |
| X56 | × | ✔(0.0135) | ✔(0.0071) | ✔ | ✔ |
| X1051 | × | ✔(0.0122) | ✔(0.005) | × | |
| X877 | × | ✔(0.0151) | × | × | |
| X512 | × | ✔(0.0019) | ✔(0.0131) | × | |
| X847 | × | × | ✔(0.01) | × | |
| X760 | × | × | × | ✔ | |
Xm represents an identifier of the genotyped SNP m in the simulated data set. Several of the significant SNPs associated with phenotype were selected across the methods meaning that they are key significant SNPs associated with complex phenotype.
× indicates either the SNP was not identified by the method. ✔ indicates the SNP was identified by the method. Numbers inside the brackets after ✔ indicate p values of the SNPs calculated using Wald test, or partial F-test.