Skip to main content

Table 3 Evaluation of the SAMS sampler

From: Characterization of a Bayesian genetic clustering algorithm based on a Dirichlet process prior and comparison among Bayesian clustering methods

  

K = 2

 

K = 4

 

K = 8

 

Program

algorithm

Microsatellite

SNP

Microsatellite

SNP

Microsatellite

SNP

DPART

Gibbs

0.016 (98)

0.292 (42)

0.228 (26)

0.476 (4)

0.315 (3)

0.523 (0)

  

2.06

1.42

3.15

2.12

5.74

3.93

 

Gibbs + SAMS

0.006 (100)

0.005 (100)

0.023 (97)

0.020 (97)

0.047 (89)

0.088 (68)

  

2.08

2.01

3.98

3.97

7.91

7.46

STRUC-

Gibbs + MC3

0.006 (100)

0.025 (96)

0.039 (91)

0.118 (67)

0.122 (52)

0.264 (20)

TURAMA

 

2.06

1.97

3.90

3.57

7.31

6.09

  1. Average , which is the partition distance between the true and inferred partition, the number of data sets in which was 0.1 or less (in parentheses) in 100 simulated data sets, and the average K values in the inferred partitions (in Italic) are shown. The number of individuals in each population was 25. The number of loci was 20 for microsatellites and 100 and for SNPs. The migration rate was 0.003 for microsatellites. MC3 indicates MCMCMC.