Phylogenetic relationships can be difficult to determine, and for this reason often require data on many characteristics, such as extensive DNA differences among species. One of the main reasons a phylogeny could be wrong is the repeated, independent evolution of a base pair or other character state by convergent, parallel, or reversed evolution.
Phylogenies are especially difficult to determine if successive branching events were closely spaced in time, because few evolutionary changes are fixed during short intervals. A related problem is incomplete lineage sorting, resulting in gene trees that differ from the species tree that one may be trying to estimate. Yet another difficulty is introgression caused by hybridization (or horizontal gene transfer).
A variety of methods are used to estimate phylogenies. The simplest is parsimony, a rule that chooses whichever phylogenetic tree requires the fewest evolutionary changes. Other methods choose among the different possible phylogenies based on their likelihoods or probabilities. Methods differ in their strengths and weaknesses, and in the kinds of data they can analyze. In many cases, different methods return very similar results.
Branching events in phylogenies can often be dated approximately, using DNA sequence differences that approximately conform to a geologically calibrated molecular clock. The rate of sequence evolution varies among parts of the genome and among clades, and can sometimes vary within clades. Tests are always necessary to confirm rate constancy. The causes of rate differences among groups of organisms are uncertain.
Phylogenies are useful for inferring histories of genes and other historical changes, such as in human cultures and languages.
An important use of phylogenies is tracing the history of evolution of characteristics, through ancestral state reconstruction. This approach has been used to synthesize ancestral DNA sequences and proteins, to better understand how their functions have evolved.
The comparative method uses convergent evolution to test hypotheses about adaptation. Statistical tools use the phylogeny to control for the effects of shared ancestry.
In modern systematics, classification of organisms is based on their phylogeny. In an ideal classification, each named taxon is monophyletic, including all the species thought to be descendants of a single common ancestor. The classification consists of nested, named monophyletic groups. Such a classification reflects evolutionary history and usually conveys a great deal of information about the species.