Table 3. Bayesian MCMC estimates based on modes of the posterior distributions and 95% credible intervals for mutation effect (e) and mutation frequency (q) parameters under two- or three-category models along with BIC relative to the model with two categories of mutation effects.

Both models include a class of mutations with zero effect on the trait.

	Parameter estimate (95% credible interval)
Model (no. mutation categories)	e1	q1	e2	q2	BIC
2	−0.031 (−0.044, −0.023)	0.042 (0.020, 0.079)	-	-	0
3	−0.024 (−0.043, −0.011)	0.071 (0.031, 0.42)	0.021 (0.010, 0.068)	0.048 (0.010, 0.41)	−147

Abbreviations: BIC, Bayesian information criterion; MCMC, Markov chain Monte Carlo.