NanoAmpli-Seq: A de novo protocol for amplicon sequencing from mixed microbial communities on the nanopore sequencing platform.

Amplicon sequencing (e.g., SSU rRNA gene sequencing), on second generation sequencing platforms leverages their unprecedented sequencing depth and significant multiplexing capacity, but are limited in their genetic resolution due to short read lengths. While third generation long read platforms can sequence entire SSU rRNA genes, their application for amplicon sequencing of conserved marker genes has been limited due to higher overall error rates (~10%). In this study, we describe the development and validation of sample preparation and data analyses approach, (i.e., NanoAmpli-Seq) that builds and expands on the previously described Intramolecular-ligated Nanopore Consensus Sequencing (INC-Seq) for profiling of complex microbial communities using full-length 16S rRNA gene sequencing. The NanoAmpli-Seq protocol involves important modifications to INC-Seq's amplicon pool preparation protocol including a significant reduction in time required for sample processing. Further, NanoAmpli-Seq includes an algorithm, i.e. chopSeq, for correction of INC-Seq consensus reads and nanoClust for read partitioning-based de novo sequence clustering and within cluster consensus calling of chopSeq corrected reads. Through these improvements, NanoAmpli-Seq accurately estimates diversity of tested mock communities and provides an average consensus sequence accuracy of 99.5±0.08% for both 2D and 1D2 sequencing chemistry on the nanopore sequencing platform. Nearly all residual errors (~0.5%) in NanoAmpli-Seq processed reads originate from deletions in homopolymer regions, indicating that homopolymer aware basecalling or error correction may allow for sequencing accuracy nearing 100%.