Table 3. Summary of multifactor dimensionality reduction (MDR) models for gene-gene interactions in breast cancer.
| No. of factors | Best candidate model | Test balance accuracy | Cross-validation consistency | P-value |
|---|---|---|---|---|
| One factor | MTHFR rs1801131 | 0.4822 | 7/10 | 0.32 |
| Two factors | MTRR rs162036 | 0.4974 | 5/10 | 0.02 |
| MTR rs1805087 | ||||
| Three factors | MTHFR rs1801131 | 0.4926 | 5/10 | <0.01 |
| MTHFR rs1801133 | ||||
| MTR rs1805087 | ||||
| Four factorsa | MTRR rs162036 | 0.5371 | 10/10 | <0.01 |
| MTR rs1805087 | ||||
| MTHFR rs1801131 | ||||
| MTHFR rs1801133 | ||||
| Five factors | MTRR rs162036 | 0.5301 | 10/10 | <0.01 |
| MTR rs1805087 | ||||
| MTHFR rs1801131 | ||||
| MTHFR rs1801133 | ||||
| MTRR rs1801394 |
aBest model predicted for breast cancer risk with highest cross-validation consistency and maximum test balance accuracy.