During a decade, overall genome related indices, including average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH), have been used as standards for the classification and identification of bacteria. On the other hand, a former study suggested that ANI-based classification is difficult for fructophilic lactic acid bacterium Apilactobacillus kunkeei. In the present study, the classification of Apilactobacillus spp., including A. kunkeei, was evaluated by multiple genome-based analyses. ANIb-based classification appropriately identified strains of Apilactobacillus spp., except for A. kunkeei-related strains. A number of strain pairings in A. kunkeei-related strains showed ANIb values around the threshold value of 95 %, based on which they were unable to be identified. On the other hand, dDDH provided clearer identification results for A. kunkeei-related strains but segmentalized them into a number of groups, while the validity of this segmentation was unclear. Certain strains shared similarities over the threshold with multiple species-level taxonomic groups in ANIb and dDDH. GTDB-Tk classifies the A. kunkeei-related strains into six species-level taxonomic groups without marked confusion, while the classification results differed from those obtained by ANIb and dDDH. The present study highlighted the inconsistent identification of A. kunkeei-related strains by the well-developed overall genome related indices, which would be a significant concern for bacterial taxonomy. Moreover, the rule adopted in GTDB-Tk, i.e., the classification of strains to taxa containing type strains showing the highest similarity, is recommended for introduction into ANIb- and GGDC-based classifications
Read full abstract