A joint research team has developed an automated design technology that enables the creation of DNA origami structures that exactly match user-drawn shapes using generative AI. The generative design model, “Generative SNUPI,” arranges DNA bases along the contour of a user-defined shape and automatically designs the bonding pathways required to assemble the structure. This breakthrough effectively enables AI to function as a nanodesigner.
The team was led by Professor Do-Nyun Kim of the Department of Mechanical Engineering at Seoul National University and Professor Chanseok Lee of the School of Biomedical Convergence Engineering at Hanyang University. The work is published in the journal Nature Communications.
The researchers demonstrated that this technology allows anyone to easily fabricate DNA origami structures in arbitrary shapes. They also confirmed that the structures can undergo shape transformations and be assembled modularly with other structures. These results suggest broad applicability in next-generation nanobio convergence technologies such as molecular robots, biosensors and drug delivery systems.
DNA origami technology is an advanced nanotechnology that folds a long single-stranded DNA molecule composed of thousands of bases into nanoscale structures of various shapes using hundreds of short DNA strands. This technique has primarily been used to fabricate regular lattice structures or polyhedral shapes for applications such as antigen-mimicking and drug delivery systems with nanometer-scale precision, as well as protein-mimicking structures with intricate geometries.
Biology demands less rigid designs
However, real biological environments—such as biomolecules and cell surfaces—often exhibit complex curved and irregular geometries, and many biological processes require dynamic structures that change shape depending on environmental conditions. Therefore, there is a growing need to fabricate DNA nanostructures with curved, freeform and reconfigurable geometries to better mimic biomolecular arrangements, enable controlled drug release and realize functional molecular robots.
Despite this need, such complex irregular structures have been difficult to design using conventional DNA origami methods. Moreover, ensuring structural stability has required iterative cycles of manual design, experimentation and refinement by experts. Although data-driven design approaches using generative AI have recently emerged across various fields, their application to DNA origami design has been limited because of the scarcity of available design and structural data.
To address these challenges, Kim’s team combined their in-house DNA origami analysis platform, SNUPI, with a diffusion-based generative AI model to develop Generative SNUPI, which automatically designs DNA origami structures that match user-defined shapes.
From contours to DNA sequences
This technology has attracted significant attention for integrating diffusion-based sampling—used to generate the three-dimensional positions of DNA bases—with a routing algorithm that designs cross-linking pathways between DNA strands, thereby enabling full automation of the design process.
Generative SNUPI arranges DNA bases along both two-dimensional and three-dimensional contours specified by the user and automatically generates the bonding pathways required for structural stability. As a result, users are provided with the DNA sequences necessary to fabricate the designed structures.
Through experiments, the research team validated that Generative SNUPI can produce a wide range of irregular and freeform nanostructures. They also demonstrated the ability to create reconfigurable structures that transition between open and closed states, as well as modular structures capable of assembly.
Lowering the barrier to design
This technology, which enables intuitive and accessible design of complex DNA origami structures, is expected to be widely adopted across various application domains. By significantly lowering the barrier to entry for DNA origami design, it reduces the time and expertise required for initial design and validation of complex nanostructures.
As a result, researchers in academia and industry will be able to rapidly explore and optimize structural candidates for diverse applications, accelerating the development of next-generation nanobio convergence technologies.
For example, Generative SNUPI could be used to design molecular robots that dynamically change shape in response to environmental conditions to regulate intracellular and extracellular transport, biosensors that selectively detect disease biomarkers, and nanocarriers that release drugs at targeted locations. In the long term, this technology is expected to contribute to advancements in precision diagnostics, personalized medicine and drug development, expanding the range of effective diagnostic and therapeutic solutions.
Furthermore, if integrated with Dark Lab automation technologies, this model could accelerate the intelligent automation of the entire DNA origami development pipeline—from design to fabrication, validation and data analysis.
Kim stated, “This study is significant in that it expands the design possibilities of complex DNA nanostructures using generative AI and reduces the difficulty of conventional design approaches that relied heavily on expert experience and manual work. Moving forward, we aim to further enhance the stability and precision of generated structures and develop this into a functional design platform applicable to real-world nanobio technologies such as biosensors, drug delivery systems and molecular robots.”
Publication details
Chien Truong-Quoc et al, De novo design of DNA origami with a generative diffusion model, Nature Communications (2026). DOI: 10.1038/s41467-026-73578-z
Key concepts
The content is provided for information purposes only.