Table 4. Summary of performance comparison for top fixed-effects models.

Methods	Comparative Power	Control of Type I Error	Control of Overlapping	Sensitivity	Specificity
ASSET	best when Ta is small	adequate	adequate	best when opposite directional effects exist	best under most settings
BE	performs well when Ta is not small	adequate	adequate	depends on m-value thresholds	depends on m-value thresholds
CPASSOC	best when opposite directional effects exist	adequate	Caution needed when traits with significant genetic correlation are analyzed	intermediate	comparable to ASSET in many settings

T_a: number of traits with non-null effects;