Abstract
Infections caused by nontuberculous mycobacteria are becoming significant due to the increasing number of vulnerable individuals worldwide. Understanding the evolutionary relationships within the genus Mycobacterium is critical for improving species identification and, consequently, enhancing diagnosis, treatment, and epidemiological tracking. Pairwise comparisons of average nucleotide identity, genome-genome distance calculations, Mash values, multilocus sequence analyses, and average amino acid identities (AAIs) revealed that the AAI metric is the best to distinguish Mycobacterium from other genera of Mycobacteriales. Furthermore, genes encoding 16S and 23S rRNAs could also be used for the genus delineation: the previously established threshold of 94.5-95.0% of the rrs was confirmed, and the value for the rrl gene was estimated at 88.5-89.0%. The genus-delineating thresholds do not confirm the proposed splitting of the Mycobacterium into five genera, and the overall performance of conserved signatures used for splitting was not satisfactory. We estimated that Mycobacterium contains at least 402 distinct species, 246 of which were identified in clinical human specimens. The obtained tree and the corresponding list of species with proposed corrections to the names made from whole-genome sequences provide a reliable framework for the identification and taxonomic positioning of novel species within the genus.