Controlled Competition Reduces Errors in Nanostructure-Based DNA Data Storage

DNA nanotechnology offers a promising route toward molecular data storage with exceptional density, longevity, and energy efficiency. Unlike sequence-based approaches that rely on large-scale synthesis and sequencing, structural DNA data storage encodes information into spatial architectures readable at the single-molecule level. However, practical implementation is hindered by substitution errors -particularly 0-to-1 “bit flips” - that arise during pooled readout and compromise data fidelity. Here, we introduce and systematically benchmark a set of thermodynamically-tuned error protection strategies for DNA ‘hard drive’ nanostructures, based on controlled competition between a label strand and a protector strand. By rationally modulating hybridisation stability through toehold truncation, internal mismatches, or self-complementary clamping, we identify a protector strand with a single cytosine-thymine (C-T) mismatch as an optimal design. This straightforward modification yields >90% readout fidelity and increases the information-theoretic channel capacity 25-fold relative to unprotected overhangs. The 1 C–T mismatch protector will enable accurate data writing and pooled readout using solid-state nanopore sensing, demonstrating scalable, high-fidelity molecular data retrieval. Overall, our findings establish new design principles for error suppression in structural DNA data storage and represent a key step toward practical, sustainable molecular information technologies.