Authors
Cai, N., Guo, W., Teng, Y., Lou, Y., Wong, S.-H., Naidu, A. S., Cona, F., Thei, F., Chen, T.-H., Bastings, M., Radenovic, A.
Abstract
Solid-state nanopores offer label-free, real-time single-molecule sensing. However, resolving fast biomolecular transport requires high-bandwidth data acquisition while the intrinsic high-frequency noise limits recovery of informative events. Here we present a hardware-software co-designed nanopore sensing platform that combines wafer-scale low-noise device engineering with deep learning-based signal reconstruction. A low-dielectric SU8 coating on silicon nitride nanopores reduces device capacitance to the pF range and suppresses high-frequency noise by up to 5-fold while maintaining facile, controllable and reproducible fabrication. This extends usable acquisition to 40 MHz and enables capture of fast molecular features. Coupled with a reconstruction model trained on synthetic translocation events embedded in experimentally measured noise, the platform recovers transient sublevels while preserving blockage edges and temporal fidelity. Using engineered DNA molecules carrying dumbbell-like barcodes, we resolve nanometer-scale structural spacings on sub-microsecond timescales, and experimentally quantify translocation dynamics within the sub-10 nanometer regime. Dual-channel measurement on a single nanopore device further demonstrates transferability of the platform by showing robust cross-channel signal reconstruction across distinct baseline noise levels. Our approach provides a general route for reliable recovery of fast event features and may enable more information-rich single-molecule sensing across diverse biomolecular targets.
Preprint server:
bioRxiv
The authors list and abstract were imported from bioRxiv on 10 Jul 2026.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 4
- Comments 0