Comparison of a long-read amplicon sequencing approach to short-read amplicons for microbiome analysis

Most microbiome studies to date rely on sequencing short amplicons of the 16S rRNA gene on Illumina's platforms. Because of the short read length, sequences often can be identified reliably only to the family or genus levels. Long-read sequencing with whole-length 16S rRNA sequencing can improve taxonomic resolution but often only to the species level. StrainID is an alternative approach that amplifies a large segment of the ribosomal operon, including the entire 16S rRNA gene, internal transcribed spacer, and a portion of the 23S rRNA gene. This longer amplicon is designed to allow ribotype-level classification. Although studies have demonstrated the utility of StrainID for several sample types, a direct comparison of StrainID to alternative approaches has not been done for saliva. Here, we compared the performance of StrainID to short-read amplicons with saliva samples as well as a synthetic mock DNA community. Short reads were amplified with primer pairs targeting the V1-V3 region of the 16S rRNA gene and were classified with several different taxonomic databases. We found that StrainID outperformed short reads not only in identifying amplicon sequence variants to the species level but also in demonstrating a key benefit with phylogenetic-based beta-diversity tests. Our results further build on establishing StrainID as a powerful method and specifically for its use with saliva samples. IMPORTANCE: The interpretation of microbiome composition studies is highly dependent on the methodologies chosen during experimental design, which affects factors such as resolution, throughput, cost, and accuracy. StrainID is an approach that can improve resolution while maintaining high-throughput and similar costs to short-read sequencing. The salivary microbiome represents a diverse community of microbes with links to a variety of health conditions and disease states. Closely related strains of bacteria can have drastically different effects on their host. Establishing StrainID as a valid approach for studying the salivary microbiome opens avenues for research that improve upon alternative methods by increasing sensitivity and accuracy compared to traditional short-read approaches.