Table 1.
Pattern of Virtual-Effect Size under Various LD Conditions
| Effect at Causal SNP | LD between Causal SNP and Marker | Virtual Effect at Marker SNP |
|---|---|---|
| β = 0 (no effect) | – | βm = 0 (no effect) |
| β > 0 | r = 1.0 (perfect LD) | βm = β (same effect) |
| β > 0 | r < 1.0 (imperfect LD) | βm < β (smaller effect) |
| β > 0 | r = 0 (no LD) | βm = 0 (no effect) |
| β > 0 | r < 0 (negative LD) | βm < 0 (effect in opposite direction) |
β is the effect size at the causal SNP, and βm is the virtual effect size observed at the marker. The LD measure is r, the Pearson correlation coefficient.