Table 1. Simulated DFEs.

	DFE type	DFE abbreviation	$S_d$	b	$p_b$	$S_b$	α
Full DFEs	Low α, low $S_d$	LALSD	−10	0.4	0.02	4	0.21
	Low α, low b	LALB	−400	0.15	0.02	4	0.21
	Low α, low $p_b$	LALPB	−400	0.4	0.005	4	0.21
	Low α, low $S_b$	LALSB	−400	0.4	0.02	0.1	0.22
	Medium α, medium $S_d$	MAMSD	−400	0.4	0.02	4	0.53
	Medium α, low $p_b$	MALPB	−400	0.65	0.005	4	0.50
	Medium α, low $S_b$	MALSB	−400	0.4	0.07	0.1	0.50
	Medium α, high $S_b$	MAHSB	−400	0.4	0.005	16	0.51
	High α, high $S_d$	HAHSD	−10,000	0.4	0.02	4	0.80
	High α, high b	HAHB	−400	0.65	0.02	4	0.80
	High α, high $p_b$	HAHPB	−400	0.4	0.07	4	0.81
	High α, high $S_b$	HAHSB	−400	0.4	0.02	16	0.81
	High $S_b$	HSB	−400	0.4	0.02	800	0.99
Deleterious DFEs	Low $S_d$	DelLSD	−10	0.4	0	—	0
	Low b	DelLB	−400	0.15	0	—	0
	Medium $S_d$	DelMSD	−400	0.4	0	—	0
	High $S_d$	DelHSD	−10,000	0.4	0	—	0
	High b	DelHB	−400	0.65	0	—	0

The DFE consists of a mixture between gamma and exponential distributions: a new mutation is deleterious with probability $1-p_b,$ and has a selection coefficient drawn from a reflected gamma distribution with mean $S_d<0$ and shape b; a new mutation is beneficial with probability $p_b,$ and has a selection coefficient drawn from an exponential distribution with mean $S_b>0.$