A Genome-Based Species Taxonomy of the Genus Complex.

There are more than 200 published species within the genus complex (LGC), the majority of which have sequenced type strain genomes available. Although genome-based species delimitation cutoffs are accepted as the gold standard by the community, these are seldom actually checked for new or already published species. In addition, the availability of genome data is revealing inconsistencies in the species-level classification of many strains. We constructed a species taxonomy for the LGC based on 2,459 publicly available genomes, using a 94% core nucleotide identity cutoff. We reconciled these species with published species and subspecies names by (i) identifying genomes of type strains and (ii) comparing 16S rRNA genes of the genomes with 16S rRNA genes of type strains. We found that genomes within the LGC could be divided into 239 species that were discontinuous and exclusive. Comparison of these species to published species led to the identification of nine sets of published species that can be merged and one species that can be split. Further, we found at least eight species that constitute new, unpublished species. Finally, we reclassified 74 genomes on the species level and identified for the first time the species of 98 genomes. Overall, the current state of LGC species taxonomy is largely consistent with genome-based species delimitation cutoffs. There are, however, exceptions that should be resolved to evolve toward a taxonomy where species share a consistent diversity in terms of sequence divergence. The genus complex is a group of bacteria that constitutes an important source of strains with medical and food applications. The number of bacterial whole-genome sequences available for this taxon has been increasing rapidly in recent years. Despite this wealth of information, the species within this group are still largely defined by older techniques. Here, we constructed a completely new species-level taxonomy for the genus complex based on ∼2,500 whole-genome sequences. As a result of this effort, we found that many genomes are not classified to their correct species, and we were able to correct these. In addition, we found that some published species are abnormally large, while others are too small. Finally, we discovered at least eight completely novel species that have not been published before. Our work will help the field to evolve toward a more meaningful and complete taxonomy, based on whole-genome sequences.