On the identifiability of substitution models
Seminar Room 2, Newton Institute Gatehouse
A prerequisite for the consistency of model-based phylogenetic methods is that the underlying model be identifiable from the joint distribution of states at the leaves.
This talk will focus on the identifiability of substitution models for tree inference, discussing some mathematical approaches to obtaining identifiability results and some analysis of their strengths and weaknesses.
In the final part of the talk, there will be some mention of the more technical details of proving the identifiability of the GTR+Gamma model, some suggestions of open problems in this area, and some indication of the direction of current work on the covarion model in progress at INI.