Nanopore-based massively parallel sensing for peptide profiling and protein identification

Abstract

Nanopore-based single-molecule sensing holds immense promise for revolutionizing proteomics, however, its practical application remains constrained by low throughput, suboptimal library preparation, and limited analytical power for complex, stochastic signals. To overcome these challenges, we develop a high-throughput nanopore sensing platform that couples a streamlined peptide library preparation strategy with an AI-driven analytical workflow, enabling accurate peptide differentiation and protein identification. Our analytical framework captures the distinct statistical signatures of peptides within massive single-molecule event streams, transforming them into reliable, information-rich fingerprints to achieve remarkable classification accuracy. We also apply this platform to establish a rapid and cost-effective workflow for antibody validation, facilitating precise epitope screening and semi-quantitative affinity determination. Critically, in a blinded study, this high-throughput sensing system demonstrates its robustness by unambiguously identifying multiple proteins from their complex enzymatic digests. By establishing an end-to-end pipeline from native proteins/peptides modification, parallel sensing to their identification, this work develops a scalable and powerful method for proteomics research. Nanopore peptide sensing faces throughput, library preparation and analysis challenges. Here, the authors develop a high-throughput platform that integrates library preparation with AI analysis to accurately identify peptides, enabling antibody validation and protein identification.