A population graph is a network structure based upon inter-stratum conditional genetic covariance (see Dyer & Nason 2004 for a more complete discussion). In this context, it is often of interest to know the statistical stability of your loci in determining the topology you see in the popgraph. Here is a way to subsample the loci you have and identify the extent to which you are asymptotically estimating a stable topology. Basically we are going to:
- Sample a subset of your loci randomly (without replacement) of a particular size (e.g., 10 loci).
- Estimate a topology.
- Measure some characteristic (or characteristics) on that topology.
- GoTo #1 a large number of times (say 100).
- Increment the number of loci being used.