Lipid nanoparticles (LNPs) are widely used for mRNA delivery and gene therapy, with their functional performance and therapeutic efficacy closely linked to nanoscale mechanical properties and formulation consistency. Nanopore-based technique offers a label-free platform to resolve LNP size, translocation dynamics, and confinement-induced deformation. Here, we investigate three LNP formulations with defined mRNA cargo loading: empty LNPs (eLNP), partially loaded LNPs containing an average of approximately 7 mRNA copies (LNP1), and fully loaded LNPs containing an average of approximately 32 mRNA copies (LNP2), using nanopipette-based resistive pulse sensing. By systematically tuning the nanopipette diameter, we demonstrate that geometric confinement critically governs sensitivity to cargo-dependent LNP deformation. Nanopipettes with 350 nm pores provide weaker confinement and limited discrimination between LNP subpopulations, whereas stronger confinement with 250 nm pores markedly enhances sensitivity to deformation. Under increased electric-field strength in the smaller nanopipettes, eLNPs exhibit pronounced reductions in relative current blockade (Δ/), consistent with enhanced deformation, whereas mRNA-loaded LNPs display progressively reduced deformability with increasing cargo content. To enable automated, multivariate discrimination of LNP populations, we implemented a multilayer perceptron neural network trained on 25 translocation-derived features. The model achieved balanced classification accuracies of approximately 81% across multiple voltage biases for the 250 nm nanopipette, reliably distinguishing three cargo formulations, eLNP, LNP1, and LNP2, based on their deformation signatures. Nanopipette-based resistive pulse sensing thus provides a single-particle approach for quantifying cargo-dependent viscoelastic heterogeneity in LNP populations and enables biophysical metrics to inform formulation screening and batch-release criteria in mRNA therapeutics. I I0
