Molecular phylogenetics is that branch of evolutionary biology which uses molecular sequence data — such as DNA, RNA, or protein sequences — to reconstruct the evolutionary relationships between organisms. When a group of related sequences is studied, it is assumed that all of them have descended from a single common ancestral sequence, and the biologist’s task is to determine the order in which these sequences diverged from one another over evolutionary time.
This relationship is represented diagrammatically in the form of a phylogenetic tree. Before applying any tree-building method (such as UPGMA, Neighbour-Joining, Maximum Parsimony, or Maximum Likelihood), it is essential to understand how a phylogenetic tree is drawn and how the information contained in it should be correctly interpreted, because a tree can be presented in several equivalent forms without any change in its underlying biological meaning.
Rooted Trees
A rooted tree is the simplest and most informative form of a phylogenetic tree because it carries a definite direction of evolutionary time.
Rooted Tree with a Time Axis
Equivalent Rooted Trees with a Time Axis
In a rooted tree drawn with a time axis, the vertical branch lengths are drawn proportional to the actual time of divergence. For example, if species A diverges from B, C, and D at 30 million years ago (Ma), species B diverges from C and D at 22 Ma, and species C and D diverge from each other at 7 Ma, this sequence of divergence events is shown clearly along the vertical axis.
- The root of the tree represents the most recent common ancestor of all the species included in the tree, and it corresponds to the earliest point of time shown on the diagram.
- The horizontal lines in such a tree carry no meaning at all — they are drawn merely to space out the species conveniently for clarity of presentation, and they do not indicate any evolutionary distance or relationship.
- Because the vertical branch lengths are scaled to time, the diagram gives an accurate visual impression of which species are evolutionarily closer and which are more distant. For instance, species C and D, having diverged only 7 Ma, appear evolutionarily close, whereas species A and D, having diverged 30 Ma, appear evolutionarily distant.
Concept of Tree Topology
A rooted tree can be visualised like a hanging mobile suspended from its root. If the horizontal branches are allowed to swing freely around the branch points (nodes), the tree can be redrawn in a completely different visual arrangement without altering the biological information it conveys. Two such differently drawn trees are said to have the same topology if the branching order of the species remains unchanged.
Rooted Phylogenetic Tree
- Topology refers strictly to the branching order of species on the tree, not to the visual left–right arrangement of the branches.
- Certain rearrangements are permissible — for example, swapping the positions of two sister species such as C and D does not change the topology, since both are still shown diverging from the same ancestral node at the same point.
- However, not all swaps are permissible. Species that are not sister groups — for example, B and C in a tree where B diverges earlier than C — cannot be swapped without altering the topology and therefore the biological meaning of the tree.
This distinction is important in examinations because two trees that look completely different on paper may, in fact, represent identical evolutionary hypotheses, and the student must be able to recognise topological equivalence rather than judging trees purely on their visual appearance.
Trees Not Drawn to Time Scale
In many practical situations, the exact timing of divergence events cannot be estimated from molecular data, although the order of divergence may still be known. In such cases, a rooted tree is drawn with branch lengths that are not proportional to time. It is therefore very important, when interpreting a published tree, to check whether the author has drawn the branches to scale; otherwise, one may form an incorrect impression of the relative evolutionary distance between species.
Relationship Between Evolutionary Distance and Time
If all the sequences under comparison are assumed to evolve at the same rate (the “molecular clock” assumption), then the evolutionary distance between any two sequences becomes directly proportional to the time elapsed since their divergence. Under this assumption, branch lengths can again be drawn to scale, with branch points labelled according to evolutionary distance instead of absolute time in millions of years. The UPGMA (Unweighted Pair Group Method with Arithmetic mean) method of tree construction is based on this assumption of a constant, equal rate of evolution across all lineages, and it produces scaled rooted trees of this type.
Rooted Tree Scaled to Evolutionary Change (Unequal Rates)
When different lineages evolve at different rates, a strict time axis cannot be maintained, since equal amounts of time may correspond to unequal amounts of molecular change in different lineages. Even so, a rooted tree can still be drawn with branch lengths scaled according to the amount of evolutionary change that has occurred along each branch.
- In such a tree, the tips of the tree (representing present-day species) do not all line up level with one another, because unequal amounts of change have accumulated along different branches.
- A longer branch leading to a particular species indicates a faster rate of molecular evolution along that lineage, while a shorter branch indicates a slower rate.
- Although the time axis is no longer strict in such a tree, useful temporal information is still retained: the tree remains rooted, so it is still possible to identify which species diverged earliest and which pair of species diverged most recently.
This type of tree is regarded as one of the most useful ways of representing phylogenetic results because it simultaneously conveys two kinds of information — the temporal order of divergence events, and the relative rates of molecular evolution among different lineages (long branches indicating high rates of change).
Unrooted Trees
Although rooted trees are the most informative, the majority of phylogenetic tree-building methods actually produce unrooted trees as their direct output.
Root Placement in Phylogenetic Trees
Features of an Unrooted Tree
- The present-day species, for which actual sequence data are available, are placed at the tips of the tree.
- The internal nodes of the tree represent hypothetical common ancestors for which no sequence data exist.
- Branch lengths in an unrooted tree are scaled according to evolutionary distance, so a branch with more change will appear longer than one with less change.
Limitations of Unrooted Trees
Unrooted trees are considerably less informative than rooted trees because they do not indicate the direction of evolutionary time. The only certainty is that internal nodes occurred earlier in time than the tips. Beyond this, an unrooted tree cannot tell us:
- Which of the internal branching events occurred earlier in time relative to another.
- Which species or lineage was the earliest to diverge from the rest.
- The direction in which evolutionary change proceeded along any branch, even though the branch length correctly tells us how much change occurred.
Because of these limitations, it is desirable, wherever possible, to convert an unrooted tree into a rooted one by correctly determining the position of the root.
Converting an Unrooted Tree into a Rooted Tree
To root an unrooted tree correctly, prior biological knowledge is required, since the sequence data alone (without an outgroup or a molecular clock assumption) cannot indicate where the root lies.
The Outgroup Method
An outgroup is defined as the earliest branching species, or group of species, in a tree — that is, the lineage known (from independent evidence) to lie outside the group of primary interest.
- Knowledge of the correct outgroup may come from the fossil record, from earlier phylogenetic studies based on morphological characters, or simply from established biological reasoning.
- In practice, it is common to deliberately include a known outgroup species along with the group under study. For example, while studying the phylogeny of Eutherian (placental) mammals, a marsupial may be included as the outgroup, since prior biological knowledge establishes that all Eutherian mammals are more closely related to one another than any of them is to the marsupial.
- Once the outgroup is identified, the root of the tree is placed on the branch that connects the outgroup to the rest of the tree.
An important consequence of this method is that different choices of outgroup produce different rooted trees, and hence imply different evolutionary relationships, even though all such rooted trees may correspond to the very same underlying unrooted tree. For example, rooting on species A, rooting on species B, or rooting on the internal branch connecting two internal nodes will each yield a differently shaped rooted tree, with different implied groupings of species — even though the branching pattern of the unrooted tree remains identical in every case. This demonstrates that the choice of root is a critical step which can materially change the biological interpretation of the tree.
The Midpoint Method
When no definite outgroup is known, the tree may be rooted using the midpoint method. This method locates a point on the tree such that the mean evolutionary distance, measured along the tree, from that point to the sequences on one side is equal to the mean distance to the sequences on the other side.
- This method essentially assumes that the average rate of evolution is the same in both halves of the tree.
- It is used as a practical alternative when independent biological evidence for identifying an outgroup is not available.
Radial versus Vertical Representation
When a tree is drawn with branches radiating outward from a central point, it is immediately clear that the tree is unrooted. However, some tree-drawing programs draw unrooted trees with all branches running vertically, which can visually resemble a rooted tree and cause confusion. It is therefore essential to check whether the author intends the tree to be read as rooted or unrooted, since some software packages, by default, root the tree arbitrarily on the first sequence listed in the data file unless the user specifies otherwise. When the true root is not known with confidence, the recommended practice is to present the tree in radial form, with species labelled around the circumference, so that no false impression of rootedness is created.
Clades (Monophyletic Groups)
A clade, also called a monophyletic group, is defined as the complete set of species that descends from a single common ancestral node, including that ancestor and all of its descendants.
- A clade must include all the descendants of a particular ancestral node — no descendant may be left out.
- For example, in a tree where species C and D share an immediate common ancestor, the group (C + D) forms a clade; similarly, the larger group (A + C + D), if they share a more distant common ancestor exclusive of other species, also forms a clade.
- A group such as (A + C), which omits some of the other descendants of their most recent common ancestor, is not a monophyletic group, because there exists no single ancestral node whose only descendants are A and C.
- Alternatively, a clade can be defined as a subset of species that are all more closely related to one another than any of them is to any other species outside that subset.
This concept is frequently tested in examinations, and students should be able to identify valid and invalid clades from a given tree diagram.
Bifurcating and Multifurcating Trees
Trees can be classified on the basis of how many branches arise from each internal node:
- Bifurcating trees — At every branch point, there is a strict two-way split. Almost all phylogenetic tree-building programs are designed to produce bifurcating trees, since it is always possible to identify the single “best” bifurcating tree according to the chosen method.
- Trifurcating / Multifurcating trees — At certain branch points, there is a three-way or many-way split rather than a simple two-way split.
A multifurcation in a tree may carry either of two possible meanings:
- It may indicate that the author genuinely believes that a rapid radiation of several lineages occurred at essentially the same point in time, such that the sequence of splitting cannot be resolved into separate bifurcating events.
- More commonly, it indicates that the author is simply uncertain about the precise order in which the branching events occurred.
A multifurcating tree therefore carries less information than a fully resolved bifurcating tree, but it has the advantage of not conveying a false sense of certainty to the reader about branching order that is not actually well supported by the data.
Reliability of Tree Branch Points
Since alternative trees, almost equally well supported by the data, may exist alongside the “best” tree produced by an analysis, it is important to have some way of assessing how reliable each branch point of a chosen tree really is. Two standard approaches used for this purpose are:
- Bootstrapping, a resampling-based statistical method.
- Posterior probabilities, typically obtained through Bayesian methods of tree construction.
Using these measures, it is often found that certain parts of a tree are very strongly supported (highly reliable), while other parts remain poorly resolved or uncertain. A degree of healthy scientific skepticism is therefore always necessary when interpreting a published phylogenetic tree, rather than accepting every branching pattern shown as an established fact.
Conclusion
A phylogenetic tree is a diagrammatic hypothesis of evolutionary relationships built from molecular sequence data. Rooted trees, whether scaled to time or to the amount of evolutionary change, provide the most complete information, since they reveal both the temporal order of divergence and the relative rates of evolution among lineages.
Unrooted trees, though more commonly produced directly by analytical methods, convey only the pattern of relatedness and the amount of change along branches, without any indication of direction in time; they must therefore be rooted using an outgroup or the midpoint method before their full evolutionary meaning can be understood.
Additional concepts such as tree topology, clades, bifurcating versus multifurcating branching, and statistical measures of branch reliability (bootstrapping and posterior probabilities) together provide the complete conceptual framework required to correctly construct, read, and critically interpret phylogenetic trees.
