Hi Hernán,
Regarding the calculation of Ne, the scheme that you propose seems to be
kind of circular: which Ne are you using for the recapitation step?
An alternative way to calculate Ne from the simulations will be from the
actual reproductive events at each time step (you can make an harmonique
mean of the Ne(t) at the end). I am not very familiar with the equations
to do this calculation in the case of overlapping generations but I
think Grimm et al (2018, doi:10.1111/2041-210X.12530) could be a good
place to look for them if you think this approach suits your needs.
Sorry, I am not really answering your first question...
For your second question, why not doing step (e) with deleterious
mutation until reach some sort of equilibrium and then doing
recapitation (if necessary) and adding neutral mutations on the tree
sequence? I do not see how your more complex approach will give
different results.
In any case, I think that if you want to start from a population with
neutral mutations, it would be more efficient and safe to simulate the
whole population with msprime >=1.0.0 (using "dtwf" model) and export
the tree sequence so it can used by SLiM as an input. This would also
allow to keep the genealogical information and that might be useful.
Miguel
On 22/09/2021 11:32,
hern....@gmail.com wrote:
> Hi all,
>
> I have two questions related to treeSeq recording and want to hear your
> opinion in case I’m violating some rules or assumptions
>
> 1) When doing *nonWF* models we often want to estimate the effective
> population size and figure out the N:Ne ratio. For this I normally run a
> long-ish simulation to achieve near equilibrium and estimate Ne=4/pi*mu.
> I want to do this step with treeseq instead, so I’m thinking:
>
> a. Calculate generation time of simulation
>
> b. Run a simulation with no mutations
>
> c. Recapitate and mutate with (mu/generation time)
>
> d. Calculate pi at time 0
>
> e. Question, should I do:
>
> i. Ne=4/pi*mu or
>
> ii. Ne=4/pi*(mu/generation time)
>
> *Do you think this is a valid approach?*
>
> 2) My second question is about doing a bit of an unconventional series
> of steps to do a *non-neutral nonWF *simulation. Neutral mutations
> normally take much longer to reach equilibrium than deleterious
> mutations, and I want to avoid running long simulations with both
> mutation types. So, I have been doing
>
> a. Run a simulation with no mutations
>
> b. Recapitate and mutate with (mu/generation time)
>
> c. Wrote a script to take the tree at time 0 and write a file in *slim
> output*
>
> d. Go back to slim (without tree recording) and read the slim output
> with neutral mutations (in theory in equilibrium)
>
> e. Include deleterious mutations from this time point and wait for those
> to reach ~equilibrium
>
> *Is this a valid approach? *I’m aware I’m losing all the genealogical
> information of the first run, but I’m not worried about that because I
> care more about what happens from step “d” onwards.
>
> cheers,
>
> Hernán
>
> --
> SLiM forward genetic simulation:
http://messerlab.org/slim/
> <
http://messerlab.org/slim/>
> ---
> You received this message because you are subscribed to the Google
> Groups "slim-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to
slim-discuss...@googlegroups.com
> <mailto:
slim-discuss...@googlegroups.com>.
> To view this discussion on the web visit
>
https://groups.google.com/d/msgid/slim-discuss/02247bd2-1a4d-4838-87cd-7a8e47fcc0a8n%40googlegroups.com
> <
https://groups.google.com/d/msgid/slim-discuss/02247bd2-1a4d-4838-87cd-7a8e47fcc0a8n%40googlegroups.com?utm_medium=email&utm_source=footer>.
--
Miguel de Navascués
UMR CBGP, INRAE
Centre de Biologie pour la Gestion des Populations
755 avenue du campus Agropolis
CS30016
34988 Montferrier-sur-Lez cedex (France)
phone:
+33499623370
fax:
+33499623345
e-mail: miguel.navascues AT
inrae.fr