Comparative analysis of human gut bacterial microbiota between shallow shotgun metagenomic sequencing and full-length 16S rDNA amplicon sequencing

Abstract

The human gut microbiome is increasingly recognized as important to health and disease, influencing immune function, metabolism, mental health, and chronic illnesses. Two widely used, cost-effective, and fast approaches for analyzing gut microbial communities are shallow shotgun metagenomic sequencing (SSMS) and full-length 16S rDNA sequencing. This study compares these methods across 43 stool samples, revealing notable differences in taxonomic and species-level detection. At the genus level, Bacteroides was most abundant in both methods, with Faecalibacterium showing similar trends but Prevotella was more abundant in full-length 16S rDNA. Genera such as Alistipes and Akkermansia were more frequently detected by full-length 16S rDNA, whereas Eubacterium and Roseburia were more prevalent in SSMS. At the species level, Faecalibacterium prausnitzii, a key indicator of gut health, was abundant across both datasets, while Bacteroides vulgatus was more frequently detected by SSMS. Species within Parabacteroides and Bacteroides were primarily detected by 16S rDNA, contrasting with higher SSMS detection of Prevotella copri and Oscillibacter valericigenes. LEfSe analysis identified 18 species (9 species in each method) with significantly different detection between methods, underscoring the impact of methodological choice on microbial diversity and abundance. Differences in classification databases, such as Ribosomal Database Project (RDP) for 16S rDNA and Kraken2 for SSMS, further highlight the influence of database selection on outcomes. These findings emphasize the importance of carefully selecting sequencing methods and bioinformatics tools in microbiome research, as each approach demonstrates unique strengths and limitations in capturing microbial diversity and relative abundances.