Table 1. Simulated Type I error rates based on 10,000 datasets.
Significance level | Optimal Omnibus-Fisher | Omnibus-Fisher | Usual Fisher |
---|---|---|---|
Independent G, M and E with binary traits | |||
0.05 | 0.0521 | 0.0516 | 0.0522 |
0.01 | 0.0100 | 0.0109 | 0.0108 |
0.001 | 0.0015 | 0.0015 | 0.0015 |
Independent G, M and E with continuous traits | |||
0.05 | 0.0483 | 0.0472 | 0.0478 |
0.01 | 0.0088 | 0.0095 | 0.0095 |
0.001 | 0.0007 | 0.0009 | 0.0009 |
G and E correlated with binary traits | |||
0.05 | 0.0488 | 0.0526 | 0.0675 |
0.01 | 0.0089 | 0.0107 | 0.0161 |
0.001 | 0.0010 | 0.0015 | 0.0029 |
G and E correlated with continuous traits | |||
0.05 | 0.0525 | 0.0509 | 0.0655 |
0.01 | 0.0093 | 0.0096 | 0.0166 |
0.001 | 0.0005 | 0.0011 | 0.0029 |