Significant impact of threshold adjustments on microbiome characterization following Nanopore sequencing

Abstract

Sequencing the full-length 16S rRNA gene is essential for improving the taxonomic resolution of bacterial identification. Long-read Oxford Nanopore Technologies (ONT) sequencing, coupled with its real-time analysis platform EPI2ME, has become increasingly competitive in microbial community profiling studies. However, the impact of threshold settings in EPI2ME Agent on the alignment-based confidence (hereafter “Alignment accuracy”) and the proportion of reads successfully assigned to a taxon (hereafter “Classification success”) has not yet been systematically evaluated. In this study, we performed ONT-based full-length 16S rRNA gene sequencing on samples collected from eggshells of wild Oriental Tits (Parus minor). By adjusting different threshold settings in the EPI2ME Agent, we examined how filtering conditions affect Alignment accuracy, Classification success, and downstream microbiome diversity analysis. Our results revealed that filtering thresholds had substantial effects on these outputs. These findings highlight the importance of carefully selecting filtering parameters in the EPI2ME Agent to balance Alignment accuracy and Classification success, and to improve the interpretability and consistency of taxonomic profiles generated from ONT sequencing data.