A Shape Annealing Approach to DNA Origami Design

The goal of this project is to automate the design of nano-sized biomechanical structures using DNA, the building block of life. These devices can be used to manufacture nanomachines, sensors, and nanorobots for use in fields ranging from biophysics to physical computing. These designs require the building of long strands of DNA that are joined in ways to provide shape and structural integrity.

Although computer aided design (CAD) tools exist to help the designer to place these strands into usable configurations and determine locations to join strands together, the designer must first fully conceive of the overall design of the structure; the CAD tools are there to fill in the detail but not generate the design layout and features themselves. This research seeks to change the way DNA nanostructures (also called “DNA origami” because of the way they are folded and joined together) are designed from automated concept generation to layout to manufacture.

Resulting design tools, to be made freely available, will enable the DNA nanotechnology and design communities to automatically generate nanostructure designs that are optimized for both form and function, and enable a broader audience of researchers in biosensing, manufacturing and robotics to apply DNA origami to their fields of study. This work broadens the advancement of design research to a new field of application, seeking efficiency and design innovation.

In addition, this work will lead to new educational endeavors, including a virtual DNA origami lab tour.

Technically, this work will merge automated design search through shape-based production grammars and stochastic search (a method called shape annealing) for the bottom-up design of DNA origami structures and their layouts. Shape grammars enable the representation of feasible and preferred DNA combinations and will be built based on the physical properties of DNA strands and their combination; simulated annealing allows for the optimal or preferred sequential build of complete DNA origami structures.

A multi-tiered shape grammar approach will be explored that begins with a macro structure and then refines the detailed features through a lower-level grammar. Providing real time feedback on the quality and progress of the design, analysis will also be multi-tiered and leverage the progression of simulated annealing from nearly random to deterministic, with pre-trained physics-based performance simulation transitioning to coarse grained molecular dynamics simulation and then to all-atom simulation.

Solutions will be compatible with current DNA origami CAD tools such as caDNAno and verified through manufacture in the wet lab.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.