evolutionary distance) can each be assigned an orthogonal axis. the tree consists only of speciation branching), more complex phylogenies may be obscured by the two-dimensional assumption.Īn obvious solution is to apply three-dimensional visualization techniques to phylogenetic trees, so that multiple independent variables (e.g.

Thus, whereas two dimensions may be adequate for 'Tree of Life' phylogenies containing exactly one sequence per species (i.e. Combining many different variables into one can make it hard to see any of them clearly. Unfortunately, trying to squeeze these very different dimensions of evolution onto a single spatial axis violates a basic principle of visual data-mining, namely that each independent variable should be assigned an orthogonal axis. Paralogous gene families, gene duplication and loss, and horizontal gene transfer are major processes of evolution that visualization should also reveal, orthogonal to speciation. However, modern gene phylogenies are far more complex than just this speciation process. For example, in one common application ('Tree of Life', analyzing a gene sequence from different species, to determine the evolutionary relationships of these species), this second dimension is used to indicate species relationships, by placing the most closely related species proximal to each other on this dimension. how distant two sequences are from their most recent common ancestor (MRCA)), the 2D assumption leaves only one dimension for representing other information. Since one dimension ordinarily represents the evolutionary distance metric (e.g. It is useful to consider how tree interpretation is limited by the assumption of two dimensionality. This 'data compression' procedure can make huge trees into simpler forms. One can simplify 2D trees by merging similar subtrees into higher taxonomy classification. Recently, there have been many efforts to improve visualization of complex phylogenetic trees. However, as sequence datasets grow exponentially due to genome and other sequencing projects, the difficulties of interpreting phylogenetic tree diagrams are also growing in size and complexity. This is not due to phylogeny being in any sense 'naturally' two-dimensional, but rather to the constraints of drawing such diagrams on paper. Traditionally, phylogenetic trees have been represented as two-dimensional (flat) diagrams. Visual representation of phylogenetic trees is an active research topic in evolutionary biology, because it is the main way that scientists view and interpret phylogenetic trees.

Freewrl zoom