Identification of microbial community in the urban environment: The concordance between conventional culture and nanopore 16S rRNA sequencing

IntroductionMicrobes in the built environment have been implicated as a source of infectious diseases. Bacterial culture is the standard method for assessing the risk of exposure to pathogens in urban environments, but this method only accounts for <1% of the diversity of bacteria. Recently, full-length 16S rRNA gene analysis using nanopore sequencing has been applied for microbial evaluations, resulting in a rise in the development of long-read taxonomic tools for species-level classification. Regarding their comparative performance, there is, however, a lack of information. MethodsHere, we aim to analyze the concordance of the microbial community in the urban environment inferred by multiple taxonomic classifiers, including ARGpore2, Emu, Kraken2/Bracken and NanoCLUST, using our 16S-nanopore dataset generated by MegaBLAST, as well as assess their abilities to identify culturable species based on the conventional culture results. ResultsAccording to our results, NanoCLUST was preferred for 16S microbial profiling because it had a high concordance of dominant species and a similar microbial profile to MegaBLAST, whereas Kraken2/Bracken, which had similar clustering results as NanoCLUST, was also desirable. Second, for culturable species identification, Emu with the highest accuracy (81.2%) and F1 score (29%) for the detection of culturable species was suggested. DiscussionIn addition to generating datasets in complex communities for future benchmarking studies, our comprehensive evaluation of the taxonomic classifiers offers recommendations for ongoing microbial community research, particularly for complex communities using nanopore 16S rRNA sequencing.

Automated summary

Microbes in the built environment have been identified as a potential source of infectious diseases. The use of long-read taxonomic tools for microbial evaluations using nanopore sequencing has shown promising results. This study analyzed the performance of different taxonomic classifiers for microbial community analysis and provided recommendations for species-level classification and culturable species identification using 16S-nanopore dataset.