Species demarcation in Bacteria and Archaea is mainly based on overall genome relatedness, which serves a framework for modern microbiology. Current practice for obtaining these measures between two strains is shifting from experimentally determined similarity obtained by DNA-DNA hybridization (DDH) to genome-sequence-based similarity. Average nucleotide identity (ANI) is a simple algorithm that mimics DDH. Like DDH, ANI values between two genome sequences may be different from each other when reciprocal calculations are compared. We compared 63 690 pairs of genome sequences and found that the differences in reciprocal ANI values are significantly high, exceeding 1 % in some cases. To resolve this problem of not being symmetrical, a new algorithm, named OrthoANI, was developed to accommodate the concept of orthology for which both genome sequences were fragmented and only orthologous fragment pairs taken into consideration for calculating nucleotide identities. OrthoANI is highly correlated with ANI (using BLASTn) and the former showed approximately 0.1 % higher values than the latter. In conclusion, OrthoANI provides a more robust and faster means of calculating average nucleotide identity for taxonomic purposes. The standalone software tools are freely available at http://www.ezbiocloud.net/sw/oat.

Thanks to the recent advancement of DNA sequencing technology, the cost and time of prokaryotic genome sequencing have been dramatically decreased. It has repeatedly been reported that genome sequencing using high-throughput next-generation sequencing is prone to contaminations due to its high depth of sequencing coverage. Although a few bioinformatics tools are available to detect potential contaminations, these have inherited limitations as they only use protein-coding genes. Here we introduce a new algorithm, called ContEst16S, to detect potential contaminations using 16S rRNA genes from genome assemblies. We screened 69 745 prokaryotic genomes from the NCBI Assembly Database using ContEst16S and found that 594 were contaminated by bacteria, human and plants. Of the predicted contaminated genomes, 8 % were not predicted by the existing protein-coding gene-based tool, implying that both methods can be complementary in the detection of contaminations. A web-based service of the algorithm is available at www.ezbiocloud.net/tools/contest16s.

The genus Turicella was proposed to harbor clinical strains isolated from middle-ear fluids of patients with otitis media. 16S rRNA phylogeny showed that it belonged to the mycolic acid-containing actinobacteria, currently classified in the order Corynebacteriales, and was closely related to the genus Corynebacterium. A new genus was proposed for the organisms as unlike corynebacteria they lacked mycolic acids and had different menaquinones. Here, we carried out large-scale comparative genomics on representative strains of the genera Corynebacterium and Turicella to check if this chemotaxonomic classification is justified. Three genes that are known to play an essential role in mycolic acid biosynthesis were absent in Turicella and two other mycolate-less Corynebacterium spp., explaining the lack of mycolic acids resulted from the deletion of genes and does not confer any phylogenetic context. Polyphasic phylogenetic analyses using 16S rRNA, bacterial core genes and genes responsible for synthesizing menaquinones unequivocally indicate that Turicella is a true member of the genus Corynebacterium. Here, we demonstrate that menaquinone and mycolic acid that have been used as critical taxonomic markers should be interpreted carefully, particularly when genome-based taxonomy is readily available. Based on the phylogenetic analysis, we propose to reclassify Turicella otitidis as Corynebacterium otitidis comb. nov.

A strictly anaerobic, Gram-stain-negative, non-spore-forming bacterium designated NSZ-14T, isolated from contaminated groundwater in Louisiana (USA), was characterized using a polyphasic approach. Strain NSZ-14T reductively dehalogenated a variety of polychlorinated aliphatic alkanes, producing ethene from 1,2-dichloroethane, propene from 1,2-dichloropropane, a mixture of cis- and trans-1,2-dichloroethene from 1,1,2,2-tetrachloroethane, vinyl chloride from 1,1,2-trichloroethane and allyl chloride (3-chloro-1-propene) from 1,2,3-trichloropropane. Formate or hydrogen could both serve as electron donors. Dechlorination occurred between pH 5.5 and 7.5 and over a temperature range of 20-37 °C. Major cellular fatty acids included C18 : 1ω9c, C14 : 0 and C16 : 0. 16S rRNA gene sequence-based phylogenetic analysis indicated that the strain clusters within the class Dehalococcoidia of the phylum Chloroflexi, most closely related to but distinct from type strains of the species Dehalogenimonas alkenigignens (97.63 % similarity) and Dehalogenimonas lykanthroporepellens (95.05 %). A complete genome sequence determined for strain NSZ-14T revealed a DNA G+C content of 53.96 mol%, which was corroborated by HPLC (54.1±0.2 mol% G+C). Genome-wide comparisons based on average nucleotide identity by orthology and estimated DNA-DNA hybridization values combined with phenotypic and chemotaxonomic traits and phylogenetic analysis indicate that strain NSZ-14T represents a novel species within the genus Dehalogenimonas, for which the name Dehalogenimonas formicexedens sp. nov. is proposed. The type strain is NSZ-14T (=HAMBI 3672T=JCM 19277T=VKM B-3058T). An emended description of Dehalogenimonas alkenigignens is also provided.

A Gram-stain negative, rod-shaped, non-spore-forming, obligate aerobic bacterial strain, JC2949 T , was isolated from grassland soil in Gwanak Mountain, Seoul, Republic of Korea. Phylogenetic analysis, based on 16S rRNA sequences, indicated that strain JC2949 T belongs to the genus Burkholderia, showing highest sequence similarities with Burkholderia grimmiae R27 T (98.8 %), Burkholderia cordobensis LMG 27620 T (98.6 %), Burkholderia jiangsuensis MP-1T T (98.6 %), Burkholderia zhejiangensis OP-1 T (98.5 %), Burkholderia humi LMG 22934 T (97.5 %), Burkholderia terrestris LMG 22937 T (97.3 %), Burkholderia telluris LMG 22936 T (97.2 %) and Burkholderia glathei ATCC 29195 T (97.0 %). The major fatty acids of strain JC2949 T were C 18 : 1 v7c, summed feature 3 (C 16 : 1 v7c and/or C 16 : 1 v6c) and C 16 : 0 . Its predominant polar lipids were phosphatidylethanolamine, diphosphatidylglycerol, phosphatidylglycerol and an unknown amino phospholipid. The dominant isoprenoid quinone was ubiquinone Q-8. The pairwise average nucleotide identity values between strain JC2949 T and the genomes of 30 other species of the genus Burkholderia ranged from 73.4-90.4 %, indicating that the isolate is a novel genomic species within this genus. Based on phenotypic and chemotaxonomic comparisons, it is clear that strain JC2949 T represents a novel species of the genus Burkholderia. We propose the name for this novel species to be Burkholderia megalochromosomata sp. nov. The type strain is JC2949 T (5KACC 17925 T 5JCM 19905 T ).