Table 2

Comparative rate distribution tests of non-synonymous and synonymous substitutions across datasets

HA: Rates free to vary

Log likelihood: -16749.63

Parameters: 229


Inferred rates for Physalis:

Inferred rates for Solanum:

dN/dS

dS

dN

Prob

dN/dS

dS

dN

Prob


2.663

1.047

2.788

0.463

1.139

0.942

1.073

0.353

1.000

0.814

0.814

0.311

1.000

2.000

2.000

0.094

0.000

0.580

0.000

0.081

0.496

0.800

0.397

0.274

0.177

1.487

0.262

0.144

0.083

0.933

0.077

0.279


a) H0: Same strength of selection

Log likelihood: -16765.49

Parameters: 228


Inferred rates for Physalis:

Inferred rates for Solanum:

dN/dS

dS

dN

Prob

dN/dS

dS

dN

Prob


1.664

1.261

2.099

0.466

1.664

0.781

1.300

0.337

1.000

0.647

0.647

0.312

1.000

2.241

2.241

0.086

0.000

0.470

0.000

0.081

0.527

0.884

0.466

0.290

0.171

1.222

0.209

0.141

0.087

1.001

0.087

0.286


Are selection strengths (dN/dS) different?

LRT = 31.722 p < 0.0001; DF = 1

b) H0: Same proportion of selected sites

Log likelihood: -16750.60

Parameters: 228


Inferred rates for Physalis:

Inferred rates for Solanum:

dN/dS

dS

dN

Prob

dN/dS

dS

dN

Prob


2.737

1.042

2.851

0.397

1.143

0.949

1.085

0.397

1.000

0.900

0.900

0.339

1.000

2.065

2.065

0.081

0.000

0.573

0.000

0.084

0.491

0.804

0.395

0.258

0.216

1.297

0.280

0.180

0.082

0.939

0.077

0.264


Are the proportions of codons under selection different?

LRT = 1.929 p < 0.165; DF = 1

c) H0: Same dN/dS and proportions

Log likelihood: -16766.96

Parameters: 228


Inferred rates for Physalis:

Inferred rates for Solanum:

dN/dS

dS

dN

Prob

dN/dS

dS

dN

Prob


1.636

1.318

2.157

0.397

1.636

0.805

1.318

0.397

1.000

0.703

0.703

0.348

1.000

2.341

2.341

0.074

0.000

0.472

0.000

0.087

0.517

0.894

0.463

0.265

0.193

1.136

0.219

0.169

0.086

1.022

0.088

0.264

Are selective regimes (dN/dS and proportions) different?

LRT = 34.647 p < 0.0001; DF = 2

d) H0: Shared distributions of rates

Log likelihood: -16764.30

Parameters: 219


Inferred joint rates:

dN/dS

dS

dN

Prob


2.507

1.034

2.593

0.189

1.000

1.139

1.139

0.338

0.543

0.797

0.433

0.251

0.086

0.988

0.085

0.222


Are the distributions different?

LRT = 29.350 p < 0.001; DF = 10


NOTE: Null models (a-d) were tested using likelihood ratio tests (LRTs) against the alternative model HA where dN and dS rates are free to vary in each dataset. Significance of p ≤ 0.05 was determined using χ2 with degrees of freedom (DF) equal to the number different parameters between models

Paape and Kohn BMC Evolutionary Biology 2011 11:243   doi:10.1186/1471-2148-11-243

Open Data