Backtracking identification techniques for predicting unclear bacterial taxonomy at species level: molecular diagnosis-based bacterial classification.

Abstract

Bacterial 16S rRNA genes are widely used to classify bacterial communities within interesting environments (e.g., plants, water, human body) because they contain nine hyper-variable regions (V1-V9) reflecting a large number of sequence variation sites between species. Short-read sequencing platform (targeting partial region of 16S rRNA gene; approximately 150-500 bp) commonly used in the 16S-based microbiome study is favored by many researchers because it is economical and can generate highthroughput sequencing data faster than long-read sequencing platforms. However, this sequencing platform has technical limitations in that it cannot clarify bacterial classification at the species level compared to long-read sequencing technology, which can cover the unclassification issue due to sequence similarity between species by targeting the 16S full-length region. In recent microbiome research-related industries, species-level high-resolution microbial classification is considered a key challenge to secure microbial resources among institutions in the field. However, the long-read sequencing platforms currently offered are still under price adjustment (demanding higher cost than short-read sequencing platforms) and have the disadvantage of low base-calling accuracy compared to short-read sequencing platforms. Therefore, this brief communication introduces the'Molecular diagnosis-based bacterial classification' technology to predict candidate species by backtracking for unclassified bacterial taxonomy at the species level in the NGS-based 16S microbiome study.