Comparative analysis of gut microbiota in children with obstructive sleep apnea: assessing the efficacy of 16S rRNA gene sequencing in metabolic function prediction based on weight status

Background Analyzing bacterial microbiomes consistently using next-generation sequencing (NGS) is challenging due to the diversity of synthetic platforms for 16S rRNA genes and their analytical pipelines. This study compares the efficacy of full-length (V1–V9 hypervariable regions) and partial-length (V3–V4 hypervariable regions) sequencing of synthetic 16S rRNA genes from human gut microbiomes, with a focus on childhood obesity. Methods In this observational and comparative study, we explored the differences between these two sequencing methods in taxonomic categorization and weight status prediction among twelve children with obstructive sleep apnea. Results The full-length NGS method by Pacbio ® identified 118 genera and 248 species in the V1–V9 regions, all with a 0% unclassified rate. In contrast, the partial-length NGS method by Illumina ® detected 142 genera (with a 39% unclassified rate) and 6 species (with a 99% unclassified rate) in the V3–V4 regions. These approaches showed marked differences in gut microbiome composition and functional predictions. The full-length method distinguished between obese and non-obese children using the Firmicutes / Bacteroidetes ratio, a known obesity marker ( p = 0.046), whereas the partial-length method was less conclusive ( p = 0.075). Additionally, out of 73 metabolic pathways identified through full-length sequencing, 35 (48%) were associated with level 1 metabolism, compared to 28 of 61 pathways (46%) identified through the partial-length method. The full-length NGS also highlighted complex associations between body mass index z-score, three bacterial species ( Bacteroides ovatus , Bifidobacterium pseudocatenulatum , and Streptococcus parasanguinis ATCC 15912), and 17 metabolic pathways. Both sequencing techniques revealed relationships between gut microbiota composition and OSA-related parameters, with full-length sequencing offering more comprehensive insights into associated metabolic pathways than the V3–V4 technique. Conclusion These findings highlight disparities in NGS-based assessments, emphasizing the value of full-length NGS with amplicon sequence variant analysis for clinical gut microbiome research. They underscore the importance of considering methodological differences in future meta-analyses.