Optimising DNA origami assembly by reducing off-target interactions

DNA origami enables the programmable self-assembly of nucleic acids into precisely defined nanostructures, yet the influence of primary base sequence on folding reliability remains incompletely understood. In particular, off-target interactions between scaffold and staple strands may introduce kinetic traps and reduce assembly yield, even when the intended Watson-Crick complementarity is preserved. Here we show that scaffold sequence strongly affects DNA origami assembly through the prevalence of off-target binding reactions implicit in the chosen base sequence. We developed a multi-objective computational framework that scores candidate scaffold sequences according to four classes of off-target interactions and selects variants predicted to minimise these effects for a given origami design. Using this approach, we identified both favourable and unfavourable scaffold regions from biological and synthetic sequences and tested them experimentally across 2D and 3D DNA origami structures. Atomic force microscopy showed that scaffolds predicted to have fewer off-target interactions consistently folded with higher yield, whereas off-target-prone scaffolds largely failed despite having fully complementary staple sets. Single-molecule optical tweezers further revealed that scaffold variants with fewer predicted off-target interactions assemble into more mechanically uniform origami structures. These results establish off-target sequence effects as a major determinant of origami folding and we provide a software tool to select scaffold sequences that minimise off-target reactions for any DNA origami design.