Table 2. Four models used for simulated data generation.
| Model 1 | bb | Bb | BB | Model 2 | bb | Bb | BB |
|---|---|---|---|---|---|---|---|
| aa | δ | δ | δ | aa | δ | δ (1+t) | δ (1+t) |
| Aa | δ | δ (1+t) | δ (1+t)2 | Aa | δ (1+t) | δ | δ |
| AA | δ | δ (1+t)2 | δ (1+t)4 | AA | δ (1+t) | δ | δ |
| Model 3 | bb | Bb | BB | Model 4 | bb | Bb | BB |
| aa | δ | δ | δ (1+t) | aa | δ | δ (1+t) | δ |
| Aa | δ | δ (1+t) | δ | Aa | δ (1+t) | δ | δ (1+t) |
| AA | δ (1+t) | δ | δ | AA | δ | δ (1+t) | δ |
δ represents the impact value of the genotype at SNP location when there is no epistasis between SNPs; t represents the change of impact value when there are interactions between SNPs.