Table 4

MLEs and likelihood ratio tests computed using Sweep_bott software for bottleneck and selective models.

Multilocus

Single-locus

mtCOI

GlobX

SAHH

PGM

northern EPR

M1 MLE

-99.718

-37.728

-11.171

-14.341

-36.475

M2 MLE

-92.660

M3 MLE

-90.460

-35.171

-9.712

-11.170

-34.407

LRT

M2-M1: 14.116

p < 0.005

M3-M1: 18.516

5.114

2.918

6.342

4.136

p < 0.025

(NS)

(NS)

p < 0.025

(NS)

M3-M2: 4.400

(NS)

southern EPR

M1 MLE

-159.367

-4.149

-57.112

-38.064

-60.042

M2 MLE

-153.211

M3 MLE

-139.745

-0.290

-52.196

-30.557

-56.704

LRT

M2-M1: 12.312

p < 0.005

M3-M1: 39.243

7.718

9.830

15.014

6.677

p < 0.005

p < 0.025

p < 0.01

p < 0.005

p < 0.005

M3-M2: 23.282

p < 0.005

M1, neutral model; M2, bottleneck model; M3, selective sweep model; NS, non-significant. Significances of the likelihood ratio tests (LRT) were tested according to a χ² table with a df equal to the difference of the number of parameters estimated for each model (df = 2, 8 and 6 for M2-M1, M3-M1 and M3-M2 multilocus LRTs, respectively, and df = 1 for the M3-M1 single-locus LRT).

Plouviez et al. BMC Evolutionary Biology 2010 10:220   doi:10.1186/1471-2148-10-220