Questions of defining demographic model for 3 populations

[Muhua Wang]

Hi Everyone,

I am working on two species arose from allopatric speciation and then have secondary contact now. I have population data for one population (pop1 in figure) that is morphologically and genetically similar to ancestral population. I also have data of two populations (pop2 & pop3 in figure) that went through the process of allopatric speciation and secondary contact. 

I tried to use dadi to define a demographic model describing this scenario. But I found some problems of defining this model using dadi. As is shown in the figure, pop2 and pop3 splitted from pop4 (which I don't have data) at different time. I tried to use "PhiManip.phi_1D_to_2D" to define pop4 splitting from pop1 and then use "PhiManip.phi_2D_to_3D" to define pop2 splitting from pop4. Then I don't know how to define the split of pop3 from pop4. I tried to use "PhiManip.remove_pop" to remove pop4 from the model. However, there are four populations existing in the model before pop3 splitting from pop4. I know dadi can just handle 3 populations. But I want to use dadi to study this scenario as it is a widely used program for demographic inference. Would you please giving me any hints or suggestions for defining the model? Thank you very much. 

Best Regrads,

Muhua

[Gutenkunst, Ryan N - (rgutenk)]

Hello Muhua,

If there's no gene flow from population 4, then it doesn't matter for the model, so you always have 3 populations. If there is gene flow from pop 4, then you have a 4 population model and dadi isn't the best tool to use. (Try SIMCOAL2.)

Best,

Ryan

--
You received this message because you are subscribed to the Google Groups "dadi-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dadi-user+...@googlegroups.com.
To post to this group, send email to dadi...@googlegroups.com.
Visit this group at http://groups.google.com/group/dadi-user.
For more options, visit https://groups.google.com/d/optout.
<Demographic_model.001.jpeg>

--

Ryan Gutenkunst

Assistant Professor
Molecular and Cellular Biology
University of Arizona
phone: (520) 626-0569, office LSS 325

http://gutengroup.mcb.arizona.edu

[Muhua Wang]

Hi Ryan,

Thank you for your suggestion. I am trying to build a model without pop4.

Best,
Muhua