Single-Molecule Analysis of SARS-CoV-2 Double-Stranded Polynucleotides Using Solid-State Nanopore with AI-Assisted Detection and Classification: Implications for Understanding Disease Severity.

Abstract

This study utilized solid-state nanopores, combined with artificial intelligence (AI), to analyze the double-stranded polynucleotides encoding angiotensin-converting enzyme 2, receptor-binding domain, and N protein, important parts of SARS-CoV-2 infection. By examining ionic current signals during DNA translocation, we revealed the dynamic interactions and structural characteristics of these nucleotide sequences and also quantified their abundance. Nanopores of sizes 3 and 10 nm were efficiently fabricated and characterized, ensuring an optimal experimental approach. Our results showed a clear relationship between DNA capture rates and concentration, proving our method's effectiveness. Notably, longer DNA sequences had higher capture rates, suggesting their importance for potential disease marker analysis. The 3 nm nanopore demonstrated superior performance in our DNA analysis. Using dwell time measurements and excluded currents, we were able to distinguish the longer DNA fragments, paving the way for a DNA length-based analysis. Overall, our research underscores the potential of nanopore technology, enhanced with AI, in analyzing COVID-19-related DNA and its implications for understanding disease severity. This provides insight into innovative diagnostic and treatment strategies.