In a typical shotgun metagenomics approach, after the DNA of an ecological community has been sequenced, it is compared to a genetic reference database of organisms with known taxonomy. Even though the number of DNA sequences and genomes in reference databases is constantly growing, there are still instances where a query sequence will not have a direct match in a reference database, and it will instead weakly align to one or more distantly related reference organisms. Furthermore, when analyzing short DNA sequences, a query DNA sequence will often match equally well to more than one reference organism, posing a challenge for its taxonomic assignation.

One solution to this problem is to apply a lowest common ancestor algorithm (LCA)(Figure 1) during taxonomic profiling to place such ambiguous assignments higher in a taxonomic tree, where they can be more confidently assigned. This idea was first implemented for metagenomics with the MEGAN program (Huson et al., 2007).