Ivan ScottiNov 17, 2022
Hello @scgm2022 , I was reading your very interesting 10.1016/J.BIOCON.2022.109704 Effective pop size paper, and
(1) I agree with you folks we have a problem with estimation of Ne
(2) I wonder whether the way we sample is the only problem we have: maybe we can test how the structure of the sample affects estimates using simulations and different sampling strategies (see pic)
Show threadSantiago C. González-Martínez

@ittocsnavi

Actually we have done something like that already, and estimates improve (as far as you don't have IBD, which would also bias the Ne estimates)

We also found that estimates of Ne based of the variance of reproductive success works fine

Personally I still think that Ne is a great genetic indicator, it is just that it is very difficult to estimate in large, continuous populations and thus not so useful from a practical point of view

Nov 17, 2022 at 10:06amWeb
Show threadIvan ScottiNov 17, 2022

@scgm2022 great, thanks! surely variance of reproductive success is "the" thing, but it's also difficult to measure with precision.

Then again, if LD-based estimates based on single sampling point scales well with the "uniform sampling" estimates, I suppose one could simply apply a multiplication factor. Do your results allow to try such an approach?

Show threadSantiago C. González-MartínezNov 17, 2022

@ittocsnavi

I am afraid they don't, as it is not clear to which surface unit the estimate refers. There are estimates of Ne for the case of metapopulations, but still they assume discret populations and not continuous ones.

Not sure there is a way out, but we are now trying to estimate Ne trends in the last 10 generations (using SNP data that is mapped) with the hope that even if the estimates themselves are biased, the trends may still be informative.

But having many doubts.

Show threadSantiago C. González-MartínezNov 17, 2022
@ittocsnavi Actually, to focus on Ne trends and not Ne itself was a suggestion of @bbrachi who I see has just migrated to Mastodon too (welcome Ben!)