Different sample size between two populations

[bio...@gmail.com]

Dear Ryan and all dadi users,

I want to infer the population dynamics (divergence time and migration rate) between two populations using dadi. The SFS was derived from angsd. In my study, one population contains four individuals and the other one contains 22 individuals. Does the uneven sample size affect the results?

I also try to fit the data with IM model, but it just shows a terrible fit. And I found the divergence time between two population is very recent and the effective population size for them are all under 100. Should I fit the data with more complex models?

Thanks

Best

Zhuqing

 

[Ryan Gutenkunst]

Hello Zhuqing,

Uneven sample size will not bias dadi inference.

From the plot, there is something seriously strange about your data. You have a huge number of variants at 50% frequency in River_IND. You should double-check your data processing pipeline. There’s no way for demography to generate such a pattern.

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 https://groups.google.com/group/dadi-user.
For more options, visit https://groups.google.com/d/optout.

[bio...@gmail.com]

Thanks Ryan. I checked the mafs.gz file, and found a lot of variants with 50% frequency in River_IND (148,938) when compared with River_ITA (56,155). This maybe the problem.

在 2018年7月11日星期三 UTC+8上午5:27:34，Ryan Gutenkunst写道：