Abstract
BACKGROUND: Genetic testing is essential for disease screening, diagnosis, prognosis, and pharmacotherapy guidance. Oxford Nanopore Technologies (ONT) offers a cost-effective platform for long-read sequencing, yet its routine use in clinical diagnostics remains under evaluation. We tested different nanopore sequencing pipelines aimed at accurately detecting single-nucleotide variants (SNV) in a gene locus spanning ⁓25 kb. METHODS: As a proof of concept, PCSK9 was selected for its relevance to cardiovascular disease and suitable sequence structure. Twelve subjects were analyzed using different sequencing flow cells, basecalling models, and SNV calling algorithms. Sanger sequencing served as the reference for performance validation. Sequencing throughput per flow cell was also estimated. RESULTS: The combination of super high accuracy (SUP) basecalling with Longshot variant calling demonstrated the highest performance across flow cells. MinION flow cell reached a perfect F1-score of 100%, while the more cost-effective Flongle flow cell remained a viable alternative (mean F1-score: 98.2% ± 4.2). Throughput analysis indicated that a single MinION flow cell could process up to 96 samples and ⁓40 long sequencing regions, whereas a Flongle flow cell could support sequencing of 96 samples and one long region. CONCLUSION: The proposed nanopore-based SNV identification workflows may support the development of long-read, targeted gene panels, offering a promising tool for both diagnostic and discovery applications, particularly in multi-gene settings such as oncology and cardiology.