A phylogenetic tree displays evolutionary relationships between taxonomical units. Nowadays, they are constructed programmatically by using different methods of inference and relying on various data sources, such as DNA and RNA sequences. The resulting evolutionary hypotheses are often contradictory, and need to be resolved by applying consensus methods. In this thesis, asymmetric median tree method (AMT) for construction of consensus phylogenetic trees is described, and along with two other approximation methods implemented in Biopython. AMT was compared to methods of strict, majority and Adams consensus trees on a number of artificial and one real data set. The comparisons were evaluated using the Robinson-Foulds and the tree resolution metrics. The results show that AMT is often the least similar to the input trees. At the same time, it is the most resolved tree, and therefore it offers the most information about evolutionary history of
taxonomical units.
|
