Table 2 Demographic simulation in fastsimcoal2 and model selection by Approximate Bayesian Computation via Random Forests (ABCRF).

From: Colonization and genetic diversification processes of Leishmania infantum in the Americas

 

Model

Pop. 1/Pop. 2

CV

Ndraws

 

Models of divergence between MT and MS L. infantum groups

AM

MT/MS

0.035

474,177

 

IMbot

MT/MS

0.086

452,533

 

IMchange

MT/MS

0.243

476,483

 

IM

MT/MS

0.085

474,263

 

SC

MT/MS

0.350

473,082

*Selected model

SCbotnomig

MT/MS

0.109

427,249

 

SCnomig

MT/MS

0.078

474,782

 

SI

MT/MS

0.014

466,136

 

PP = 0.504

    

MIGMT>>>MS = 0.254

    

MIGMS>>>MT = 0.300

    

Models of divergence between MT and OW L. infantum groups

AMbot

MT/OW

0.304

432,323

*Selected model

AM

MT/OW

0.186

458,125

 

IMchange

MT/OW

0.106

470,330

 

IM

MT/OW

0.215

459,566

 

SC

MT/OW

0.161

421,405

 

SCnomig

MT/OW

0.013

464,907

 

SIbot

MT/OW

0.003

385,170

 

SI

MT/OW

0.012

409,244

 

PP = 0.485

    

FOU = 0.204

    

Models of divergence between MS and OW L. infantum groups

AMbot

MS/OW

0.385

413,704

*Selected model

AM

MS/OW

0.161

472,457

 

IMchange

MS/OW

0.145

473,388

 

IM

MS/OW

0.170

471,073

 

SC

MS/OW

0.025

471,677

 

SCnomig

MS/OW

0.031

472,251

 

SIbot

MS/OW

0.035

463,789

 

SI

MS/OW

0.048

457,084

 

PP = 0.521

    

FOU = 0.292

    
  1. CV classification vote, i.e., the number of times a model is selected in a forest of 1000 trees (the model with the most votes corresponds to the model best suited to the dataset), FOU bottleneck size, i.e., the fraction of prior population size at the end of the bottleneck, MIGx»>y migration rate from x to y, MS Mato Grosso do Sul, MT Mato Grosso, Ndraws number of parameter draws simulated by fastsimcoal2 as input for ABCRF, OW Old World, Pop. population, PP ABCRF approximation of the posterior probability of the selected model.
  2. In fastsimcoal2 simulation, values for past and present population sizes were drawn randomly from a uniform distribution between 100 and 106 individuals. Values for time of secondary contact were drawn randomly from a uniform distribution between 0 and 2 × 104 generations before present. Values for relative migration rates between populations were drawn randomly from a log-uniform distribution between 10−10 and 0.1. Values for bottleneck size were drawn randomly from a uniform distribution between 0.05 and 0.5. The mutation rate was fixed at 1.99 × 10−9 mutations per bp on all chromosomes. The ten different demographic models are illustrated in Supplementary Fig. 8 and template file content is provided is provided in Supplementary Data 7 at Zenodo36.