Center for Scientific Workshops in All Disciplines

Current Workshop | Overview | Back | Home | Search | | ||||||||||

## The Future of Phylogenetic Networks |

Biologists continue working with phylogenetic trees, even if
they realise that it is not the adequate model for
the genetic phenomena they study; phylogenetic trees are less suited to
capturing reticulate evolutionary phenomena such as hybridisation,
recombination or lateral gene transfer. Phylogenetic networks are a generalisation
of phylogenetic trees that can display such complex evolutionary scenarios; the
long-term goal of the workshop is to develop practical algorithms for these
networks. This will allow the current tree model, which incorporates only
vertical descent from parent to offspring, to be expanded to include horizontal
(reticulate) evolution. The first purpose of the workshop is that biologists from
various sub-disciplines explain to math-ematicians
and computer scientists their specific requirements for phylogenetic network
models, in order to make them useful for their work. In return, the mathematicians
will outline the inherent mathematical limitations of various algorithmic
approaches, and highlight potentially fruitful avenues for future development.
By interacting in lectures and group discussions, the key features of phylogenetic
networks are identified that are important for any successful algorithm, both
from the biological and computational aspects. The ultimate goal of the workshop is that, through close
collaborations, mathematical research on phylogenetic networks will lead to
more practical methods, ultimately resulting in more widespread use of
phylogenetic network software by biologists.
Although the workshop was not aimed at direct new mathematical
or biological results, some subgroups of people, especially mathematicians and
computer scientists, have made progress on subjects related to the workshop.
The best learning moments of the workshop have been for the math/cs people to see the various ways that biologists are
interested in using networks i.o.trees, which each
require their own specific model. For the biologists it was to see where the mathematical
developments are currently: what they may expect and what not, currently. Also they
realised that they will need to make some effort to
define their problems more clearly. Given the goals of the workshop, it was an
enormous success, and in fact a success beyond expectation.
The organisation of the workshop
appeared to work out exceptionally well. Two aspects triggered the success.
Firstly, the biologists had been selected carefully; all of them feeling the
need for networks as an alternative to trees, and being interested to explain
why they still were not using existing software/algorithms. Next to that David
Morrison, as the biology organiser of the committee
played an excellent role in moderating the discussions whenever necessary. This
made the discussion sessions extremely lively and informative until the very
last moment of the workshop. Secondly, all lecturers were given very precise instructions
about what was expected from them. Especially the math/cs
researchers were asked to give overview lectures with a lot of empathy for the
biologists in the audience. This worked out very well, maybe not in the least
because we were lucky - or less moderately, some members of the organising committee appear to have enough reputation
worldwide - that almost all prominent international math/cs/bioinformatics
researchers in the field came to the workshop, even from the remotest places on
earth like New Zealand. Thus, the knowledge on the actual state of the art on
the algorithmic side of the subject was present, and they were the perfect
group of people to expose the overviews. At the end of the workshop everybody agreed that we should try
to actively keep the momentum going that has been created. For sure, we will
actively visit, correspond through and contribute to the blog on phylogenetic
networks. We will investigate possibilities to write a
EU grant proposal to help us to realise our goals:
creating software ready for biologists to help them in their studies and on the
way solving a lot of beautiful mathematical challenges. [Back] |