Parts | Days | Selection | Search | Updates | Downloads | Help
BP: Fachverband Biologische Physik
BP 7: Posters: Statistical Physics in Biological Systems
BP 7.15: Poster
Monday, March 14, 2011, 17:15–20:00, P3
Supertree construction using superparamagnetic clustering — •Pascal Fieth1, Alexander K. Hartmann1, and Olaf R.P. Bininda-Emonds2 — 1Institute of Physics, University of Oldenburg — 2Department of Biology and Environmental Sciences, University of Oldenburg
Superparamagnetic clustering [1] is a non-parametric clustering method for a set of data points using the Potts model with a suitable distance definition. In a numerical simulation [2] of the data points represented as spins in a thermodynamic system, regions of aligned spins, corresponding to clusters, are detected. Here, this method is used for supertree construction, a phylogenetic approach to merge phylogenetic trees only according to their tree topologies, so that overlapping rather than identical taxon sets are needed [3].
The corresponding distance definition is empirically optimized for the used method. Further the dependency of the quality of the reconstruction of a known tree on the size and number of the used source trees is analyzed.
[1] M. Blatt, S. Wiseman and E. Domany, Superparamagnetic Clustering of Data, (Physical Review E, 1998)
[2] A.K. Hartmann, Practical Guide to Computer Simulations, (World Scientific, 2009)
[3] O.R.P. Bininda-Emonds The evolution of supertrees, (TRENDS in Ecology and Evolution, 2004)