Systematics and Evolutionary Biology

Supertrees and phyloinformatics

The molecular revolution combined with the availability of sequence data in public databases such as GenBank, Swiss-Prot, or HOVERGEN has increased the scope of phylogenetic inference tremendously. Never before has so much informative data been available. However, the wave of new data has also brought a wave of new questions and problems: how can we best analyze the data so as to achieve the most comprehensive and robust phylogenies possible?

Supertree construction

Phylogenetic supertrees represent an alternative approach to traditional phylogenetic methods where the trees obtained from primary character data sets are combined instead of the data sets themselves. It therefore potentially makes use of a greater part of the historical and global systematic database because even incompatible data types (e.g., DNA hybridization data and DNA sequence data) can be combined. Supertrees have been used to successfully derive large, comprehensive phylogenies for many groups. Using supertrees, we were able to generate the first virtually complete species-level phylogeny of extant mammals. The mammal supertree, like the other comprehensive supertrees, has proven to be invaluable for understanding the biology of the group.

The mammal supertree from Bininda-Emonds et al. (2007).

Click here for an expanded version.

The mammal supertree

Research into supertrees in the group is continuing on both empirical and theoretical fronts:

  • Empirical
    • an update of the carnivore supertree (Katrin Dix)
    • a supertree of snakes (Lars Schlegel)
  • Theoretical
    • further investigations into determining nodal support for supertrees
    • continued development of supertree algorithms and methods
    • merging of the supertree and supermatrix approaches in a divide-and-conquer framework
    • development of computer programs
Some useful links for supertrees include:

Additional project members: Olaf Bininda-Emonds
External collaborators: Vincent Berry (Université Montpellier II), Andy Purvis (Imperial College London), Charles Semple (University of Canterbury)




Apart from supertrees, which really do fall into phyloinformatics as well, we are also investigating other aspects of this field, often using supertrees to achieve a more comprehensive framework than might otherwise be achievable:

  • developing tools for the automated handling of DNA sequence data, including:
  • comparative investigation of:
    • rates of molecular evolution across mammals and genes (using
    • phylogenomic conflict between genes and genomes within mammals
    • chromosomal evolution in vertebrates
  • investigations into limitations of and new approaches for building really large phylogenies, including:
    • building really large phylogenies
    • designing objective stopping criteria for the non-parametric bootstrap
  • building molecular phylogenies!

Rates of molecular evolution in mammals

Comparative rates of molecular evolution in mammals. Red clades show either significantly higher (>0) or lower (<0) relative rates of molecular evolution than the mammal average. (Adapted from Bininda-Emonds, 2007)
Interesting phyloinformatic resources include:
Project members: Olaf Bininda-Emonds
External collaborators: Vadim Goremykin (IASMA Research Center), Michael Ott (TU München), Alexandros Stamatakis (LMU München)