Covalently functionalized MoS2 nanopore with amino acids for label-free protein residue discrimination.

Precise detection of amino acids is critical for proteomics and molecular recognition. Nanopore sensing enables label-free single-molecule analysis but often lacks sufficient spatial resolution to control translocation dynamics and resolve protein structure function relationships. Here, we employ all-atom Molecular Dynamics simulations to investigate Cysteine-functionalized MoS2 nanopores as stochastic sensing elements. Cysteine modification enhances peptide-nanopore interactions, improving sensitivity and discrimination, while linkage to polar or nonpolar residues allows tunable sensing performance. Hydrogen bonds modulate translocation through a gating effect, with shorter lifetimes than van der Waals interactions and higher voltage sensitivity. Integrating machine learning with a seven-dimensional feature matrix, the system accurately distinguishes up to 16 amino acids and multiple post-translational modifications, including phosphorylation, glycosylation, methylation, and acetylation. This strategy provides a general platform for rapid, label-free single-molecule protein detection and offers theoretical guidance for designing bioinspired nanopores.