Estimating current Ne population

[AudeCaizergues]

Dear Ekaterina,

I as able to run GADMA successfully and even if the generated plot give me a global estimate of the current Ne of my population, I would like to have access to the exact number. However, I checked in the manual and I didn't find any explanation on how to use my estimates to compute the current Ne values (outside of the plot).

For instance here, my best model gave the following parameters:

Run 5 -2700.30 [Nanc = 617187] [ [ 1 pop split 3.67% (s1) [0.037(s1*_Nanc_size), 0.963((1-s1)*_Nanc_size)] ], [ 1000.0(t1), [61718746.147(nu11), 24096350.787(nu12)], [[0, 0.00e+00(m1_12)], [4.93e-10(m1_21), 0]], [Exp(dyn11), Sud(dyn12)] ] ] f (theta = 32419.38)

And µ and L were :

mu = 1.2e-08
L = 1094325

Thank you for your help,

Aude

[noscode]

Dear Aude,

Thank you for using GADMA! I am glad to help you with your question.

What do you mean by current Ne? Modern size of population or an ancestral size? The answer for any of those is in the line you took from GADMA output:
Run 5 -2700.30 [Nanc = 617187] [ [ 1 pop split 3.67% (s1) [0.037(s1*_Nanc_size), 0.963((1-s1)*_Nanc_size)] ], [ 1000.0(t1), [61718746.147(nu11), 24096350.787(nu12)], [[0, 0.00e+00(m1_12)], [4.93e-10(m1_21), 0]], [Exp(dyn11), Sud(dyn12)] ] ] f (theta = 32419.38)

This line contains the exact values for inferred demographic parameters. The description of this output could be found here.

From this output the size of the ancestral population is equal to 617,187 individuals. This population was divided into two subpopulations 1,000 generations ago (or years if you have set time per generation for GADMA). The first subpopulation experienced exponential growth from 617,187 * 0.037 up to 61,718,746 individuals and the second had a constant size of 24,096,350.

Does it answer your question or you meant something else?

I also see the value of 61,718,746 for the nu11 parameter, it is equal exactly to 100*Nanc which indicates its upper bound. Usually, when a parameter hits the bound it is better to change the bound and find a more correct value of the parameter. I would recommend to change the upper bound for population size by setting in params_file of GADMA:
maxN: 150 or 200

Best regards,
Ekaterina

[AudeCaizergues]

Dear Ekaterina,

Thank you for your answer, I didn't realise that nu11 and nu12 were the final population sizes, that's exactly what I need !
Thank you for the heads up on the 100*Nanc value, I will re-run the model changing the maxN to see if it changes.

Cheers,

Aude

[noscode]

Happy to help!

[AudeCaizergues]

Dear Ekaterina,

I'm opening this issue again to check something related to the topic.
Since I used a model where my parameters are not in genetic units, what is the unit of the migration rate ? Number of individual that migrate from pop1 to pop2 per generation ?
In the manual I read "GADMA shows model parameters in genetic units. To scale them from one should multiply migration rates by 2 only. Other units are as usual. In case when option Relative parameters is set to True one should first rescale from units of Nref: sizes of populations - multiply by Nref, time - multiply by 2 * Nref migration rates - divide by 2 * Nref." But nothing is specifyed for when results are not in genetic or relative units.
Thank you for your help,

Aude

[noscode]

Dear Aude,

You are almost correct, the migration rate in physical units corresponds to the proportion of chromosomes per generation in a
population that are new migrants from population.

Thank you for the feedback about the documentation, I will add information there.

Best regards,
Ekaterina