For candidate variants, discovery power is adequate for any CMV reactivation in the allelic and dominant genetic models but not in the recessive model, while discovery power is limited for high-level CMV activation and disease. Plots show the percentage of tested recipient variants with post-hoc power at the level on the x-axis with each of the 3 genetic models and an assumed clinically significant HR of 1.5 or 0.67, based on the observed discovery standard error and a .01 threshold of statistical significance for (A) CMV reactivation, (B) high-level CMV reactivation, and (C) CMV disease. Results for donor variants are essentially the same.