Let's imagine a scenario in which I have 100 individuals from five different species of the same genus. Their phylogenetic relationship is known (a tree is available). I want to look at the overall correlation between two of their traits at the individual level while accounting for the fact that they belong to different species. Is it possible to do so? A mixed effect model (MEM) with species as a random effect does the job but it estimates a covariance matrix from the data rather than from a phylogenetic tree. Phylogenetic Generalized Least Squares (PGLS) regression allows to specify different ways in which the phylogenetic structure is expected to affect the covariance in trait values, but PGLS focuses on taxa rather than individuals. The R package MCMCglmm allows to specify individuals as a random factor, but its focus is still at the species level as far as I understand.
How to perform a “hybrid” analysis between PGLS and MEM, in that it focuses at the individual level rather than the species level but it estimates the covariance matrix from a phylogenetic tree? Is such an analytical tool even available (ideally in R)?
