# Machine Learning‐Driven Nanopore Sensing for Quantitative, Label‐Free miRNA Detection

## TL;DR

This paper introduces a machine learning method using nanopore sensors to detect microRNAs with high accuracy and sensitivity for potential disease diagnostics.

## Contribution

The study introduces a CNN-based analytical framework that outperforms traditional methods in classifying nanopore sensing signals for miRNA detection.

## Key findings

- The CNN model achieved near-perfect classification performance (accuracy = 0.99, precision = 0.99, recall = 0.99) for delayed versus non-delayed events.
- Nanopore-derived delay metrics closely matched RT-qPCR validation data, confirming the method's reliability.
- The CNN model demonstrated superior sensitivity and robustness compared to moving standard deviation and spectral entropy methods.

## Abstract

Nanopore sensors offer exceptional sensitivity for detecting single molecules, making them ideal for early disease diagnostics. In this study, we present a multiplexed nanopore‐based assay that combines DNA‐barcoded probes with advanced computational analysis to detect microRNAs (miRNAs) with high specificity and accuracy. Each probe selectively binds its target biomarker and induces a characteristic delay in the ionic current signal upon translocation through the nanopore. We evaluated three analytical strategies for classifying delayed versus non‐delayed events: (1) moving standard deviation (MSD), (2) spectral entropy (SE), and (3) a convolutional neural network (CNN). While MSD and SE rely on manually defined thresholds and exhibit limited sensitivity, the CNN model, trained on image representations of raw current traces, achieved near‐perfect classification performance across all metrics (accuracy = 0.99, precision = 0.99, recall = 0.99). Grad‐CAM visualization confirmed that the CNN model focused on relevant signal regions, enhancing interpretability and generalizability. All methods produced sigmoidal concentration‐response curves consistent with expected binding kinetics, and nanopore‐derived delay metrics closely matched RT‐qPCR validation data. All three methods were capable of distinguishing between signal classes; however, the CNN model demonstrated superior sensitivity and robustness. This work highlights the importance of data interpretation in nanopore sensing and presents a comparative framework for binary event classification. The findings pave the way for the development of machine learning‐driven nanopore diagnostics capable of detecting diverse biomarker types at the single‐molecule level.

Nanopore sensors combined with DNA‐barcoded probes enable detection of multiple biomarkers in parallel at the single‐molecule level through characteristic delays in ionic current signals. Three analytical strategies are compared: moving standard deviation, spectral entropy, and a convolutional neural network (CNN). The CNN achieves superior sensitivity and robustness, offering a scalable, automated framework for multiplexed nanopore diagnostics.