Modelling Phylogenetic Signals Using Eigenvector Maps
Computation of Phylognetic Eigenvector Maps (PEM) and simulation of trait evolution among phylogenetic trees or any directed graph representing reticulated phylogenies.
PEM are sets of orthogonal basis vectors (eigenvectors) that are tailored to represent trait evolving neutrally (i.e., with trait values changing smoothly along the edges and nodes of the phylogenies) or non-neutrally (i.e., with trait values shifting abruptly at the nodes of the phylogeny following adaptive shifts related with changing environment or niche). Each basis (eigen) vector represents a potential trait evolution pattern and linearly combining sets of such patterns enables one to represent trait (or meta-trait) evolution (phylogenetic modelling) or partial out phylogenetic variation when assessing ecological hypothesis (e.g., testing functional trait correlation hypotheses).
MPSEM also features functions to simulate trait evolving neutrally or non-neutrally, with specified optimal trait values (and shifts thereof) and evolution rates, and along either “classical” phylogenetic trees (i.e., not involving lateral gene transfer) or reticulated phylogenies represented by a directed graph (i.e., involving lateral gene transfer through, e.g., hybridization). That functionality is useful to simulate the range of trait values that can emerge from different trait evolution scenario.
Maintained by Guillaume Guénard – Université de Montréal