Optimizing Microbiome Reference Databases with PacBio Full-Length 16S rRNA Sequencing for Enhanced Taxonomic Classification and Biomarker Discovery

This study aimed to determine if full-length 16S rRNA sequencing data produced by PacBio could be used to optimize reference DBs and be applied to Illumina V3-V4 targeted sequencing data for microbial study. Methods: Oral and gut microbiome data (PRJNA1049979) were retrieved from NCBI. DADA2 was applied to full-length 16S rRNA PacBio data to obtain amplicon sequencing variants (ASVs). The RDP reference DB was used to assign the ASVs, which were then used as a reference DB to train the classifier. QIIME2 was used for V3-V4 targeted Illumina data analysis. BLAST was used to analyze alignment statistics. Linear discriminant analysis Effect Size (LEfSe) was employed for discriminant analysis. Representative codes and workflow is presented.