Skip to main content

Table 2 Results of constrained topology tests using Bayesian and maximum likelihood methods

From: Dynamic evolution of the GnRH receptor gene family in vertebrates

Model

Constraint

Topology test

  

Bayes factor

AU test

SH test

Hyp 1: Type I GnRH receptors are not monophyletic

Negative

15.300**

N/A

N/A

Hyp 1a: Chimaera Type I receptor is within the monophyletic Type II subfamily

Positive

N/A

> 0.001

0.013

Hyp 1b: Chimaera Type I receptor is ancestral to vertebrate receptors

Positive

N/A

0.002

0.013

Hyp 2a: Type IIb receptors are not monophyletic

Negative

12.970**

N/A

N/A

Hyp 2b: Type IIa-2 receptors are not monophyletic

Negative

13.280**

N/A

N/A

Hyp 2c: Type IIa-3 receptors are not monophyletic

Negative

15.300**

N/A

N/A

Hyp 3a: Type IIa-1 and IIb subfamilies are monophyletic to the exclusion of all other clades

Positive

2.080**

> 0.001

> 0.001

Hyp 3b: Type IIa-1 and IIa-2 subfamilies are monophyletic to the exclusion of all other clades

Positive

(-5.660)**

0.387

0.846

Hyp 3c: Type IIa-1 and IIa-3 subfamilies are monophyletic to the exclusion of all other clades

Positive

2.000**

0.008

0.127

Hyp 4: Type IIa-3 and IIb subfamilies are monophyletic to the exclusion of all other clades

Positive

4.070 *

> 0.001

> 0.001

Hyp 5: Tunicate and vertebrate receptors are monophyletic to the exclusion of all other clades

Positive

30.900**

0.624

0.968

  1. Hypotheses tested using topological constraint tests, which are also illustrated in Figure 7. Hypotheses were compared to the null hypothesis that vertebrate receptors are monophyletic. Specified clades for each test were either constrained to be monophyletic (positive constraint) or not allowed to be monophyletic (negative constraint; limited to Bayesian analyses). Significant results indicate rejection of the hypothesis if the values are positive and acceptance if the values are negative. Bayes Factor scores result from comparisons to a null model with a positive constraint on the vertebrate branch, using a stepping stone model [37]. The strength of the inference is * = strong or ** = very strong [40]; AU = p-values for the Approximately Unbiased test [38]; SH = p-values for the Shimodaira-Hasegawa test [39].