Read-Level Error Characterization of Rolling-Circle Amplification-Based Nanopore Sequencing of the Circular DNA Virome

Oxford Nanopore technology enables cost-effective, portable, long-read analyses of pathogen genomes. Accurate detection and interpretation of small circular viral genomes, including Anelloviridae, remain challenging due to limited base-level error quantification in rolling-circle amplification (RCA)-derived datasets. Here, we characterized read-level sequencing error profiles using M13mp18, a 7.2 kb circular phage genome, subjected to 1X and 3X shearing during library preparation. M13mp18 DNA was serially diluted into pooled anellovirus-positive plasma DNA extracts. Using custom error-analysis pipelines, we quantified mismatch, insertion, and deletion rates and evaluated consensus reconstruction accuracy across simulated sequencing depths. Since metagenomic viromes contain mixtures of related genomes and uneven coverage across taxa, depth-normalized subsampling was used to assess the precision of read-level error estimates under heterogeneous coverage. Across four benchmarked datasets, per-base error rates ranged from 0.018 to 0.022 errors per aligned base. Complete M13mp18 reference reconstruction was achieved at input levels ≥ 4.6 log10 copies, and consensus sequences reached 100% identity at depths ≥ 15X when sufficient reads were available. Below 4.6 log10 input copies, recovery was inconsistent. These findings provide a controlled empirical characterization of read-level error behavior in RCA-derived nanopore sequencing and support the interpretation of circular DNA virome data generated in complex metagenomic backgrounds.