Phylogenetic trees for prokaryotes

(See Edgar 2018, Taxonomy annotation and guide tree errors in 16S rRNA databases for more details and references).

The history of vertical (i.e., clonal or sexual) inheritance can be represented as a binary tree. Inheritance can also occur by horizontal transfer of DNA between cells, which has been inferred to occur at a vast range of evolutionary distances ranging from members of the same species to exchanges between different kingdoms (Jain et al., 2002).

If a trait is acquired by horizontal transfer, then a graph representing its inheritance is not a tree. If horizontal transfer is rare, then in principle a phylogenetic tree could be constructed by following only vertically inherited traits.

Gene content is sometimes highly variable between sets of prokaryotic genomes with otherwise highly similar sequences, which has been interpreted to imply that the optional genes which may be absent are frequently horizontally transferred (Gogarten and Townsend, 2005). If traits used for classification are acquired by horizontal transfer, then it may be impossible to achieve consistency between taxonomy and phylogeny.

It is widely believed that the 16S rRNA gene is unlikely to be horizontally transferred and is thus well-suited as a phylogenetic marker (Woese, 1987), though recent evidence suggests that transfer of 16S rRNA genes can occur when sequence identity is as low as ~80% (Kitahara and Miyazaki, 2013).

The evolutionary history of a single gene can be represented as a binary tree if its sequences are inherited as intact units, i.e. providing that rearrangement events such as cross-over recombinations do not occur.

If rearrangement events and horizontal transfers are rare for the 16S rRNA gene, as is widely believed to be the case, then its true gene tree is likely to be a good approximation to the true phylogenetic tree based on vertically inherited traits, assuming that the latter tree can be meaningfully defined.

However, even if there is good reason to believe that the true gene tree would be an accurate guide to taxonomy, it is very challenging to estimate the gene tree from multiple sequence alignments that span vast evolutionary distances, and in practice the trees almost certainly have pervasive branching order errors.