Robust Phylogeny Estimation Based on Profile Distances: A Study of the Chlorophyceae
(Chlorophyta)
Project leader: Matthias Wolf (PI), Tobias Müller (Co-PI), Thomas Dandekar (Co-PI)
Staff: Joachim Friedrich (diploma thesis)
Objectives: The three phylogenetic projects are given in more detail to illustrate differences in their objectives. This could also be applied to other areas in the department, but was not attempted to save space and avoid too many projects. In phylogenetic analysis we face the problem that several subclade topologies are known or easily inferred and well supported by bootstrap analysis, but basal branching patterns cannot be unambiguously estimated by the usual methods (maximum parsimony, neighbor-joining, or maximum-likelihood, nor are they well supported.
Approach: We represent each subclade by a sequence profile and estimate evolutionary distances between profiles to obtain a matrix of distances between subclades. Our estimator of profile distances generalizes the maximum likelihood estimator of sequence distances. The basal branching pattern can be estimated by any distance-based method, such as neighbor-joining. Our method then inherits the robustness of profiles and the time efficiency of neighbor-joining.
Progress: Phylogenetic analysis of Chlorophyceae with traditional methods (MP, NJ, ML, and MrBayes) reveals seven well supported subclades, but the methods disagree on the basal branching pattern. The tree reconstructed by our method (PNJ, profile neighbor-joining) is highly supported and can be confirmed by known morphological characters.
Significance: The proposed method helps to solve a general problem in phylogeny. It is applicable to all trees with low bootstrap support on the basal branching pattern. From a computational point of view, there are some promising approaches. For example, we may use the concept of profile distance in a divide-and-conquer algorithm that iteratively applies neighbor-joining on growing well supported subclades. In this way we may obtain better support on basal branching patterns in supertrees, e.g., in the tree of life or within the crown eukaryotes.
Future projects: (1) A diploma thesis by Joachim Friedrich entitled: ProfDist: A tool for the construction of large phylogenetic trees based on profile distances, and (2) application of external funding for a PhD student evaluating and optimizing PNJ and ProfDist.
Collaborations:
Prof. Peter Rossmanith (RWTH Aachen)
Dipl. Biol. Sven Rahmann (FU Berlin)
Publications:
Current external funding: None (new project)