FlexTax: Assigning species-level taxonomy to full-length 16S rRNA gene sequences

FlexTax is a hybrid bioinformatics pipeline that we developed to improve species-level taxonomic classification of full-length 16S rRNA gene amplicon sequences generated using long-read sequencing technologies such as the PacBio Sequel IIe. The pipeline combines two complementary classification algorithms, QIIME2's Naive Bayes sk-learn classifier and the alignment-based VSEARCH, to enhance taxonomic resolution and accuracy. FlexTax also includes steps for HiFi read conversion, adapter trimming, and denoising using DADA2. The pipeline supports a wide range of sample types and sequencing conditions and provides a framework for microbial community analysis at species-level resolution while addressing key limitations of existing pipelines. We tested the performance of our dual classification approach on full-length 16S rRNA gene sequences obtained from canine and feline fecal and oral samples.