Post date: Mar 08, 2020 11:12:14 PM
Spatial versus genetic distance
What I did:
Use Euclidian distance to estimate pairwise genetic and spatial distance between individual ramets.
- The spatial data = coordinate (lat and long) for each ramet
- The genetic data = probability of being heterozygote (as calculated based on Phred score) for each individual for each SNP
For each comparison of two individuals, I thus have a value for the genetic distance and a value for the spatial distance. Euclidian distance is a starting point and not the best distance to be used in both cases :
- use metric distance for spatial data?
- look into genetic distance calculations
I was expecting a linear relationship : the further you are from a tree, the most genetically different. It does not seem to be the case from the above quick graph. The points distribution suggest noisy genetic repartition (scenario where roots expanded over old ones and blurry the genetic spread and history) However, results are to be handled with care because of the reason stated above (maybe not the most relevant distance measures to be used)