DNA based self-assembly and nano-device: Theory and practice

by Yin, Peng, Ph.D., Duke University, 2005, 194 pages; AAT 3181510



Abstract (Summary)

The construction of complex systems at the 1-100 nanometer (1 nanometer = 10 -9 meter) scale is a key challenge in current nanoscience. This challenge can be most effectively addressed by the "bottom-up" nano-construction methodology based on self-assembly, a process in which substructures autonomously associate with each other to form superstructures driven by the selective affinity of the substructures. DNA, with its immense information encoding capacity and well defined Watson-Crick complementarity, has recently emerged as an excellent material for constructing self-assembled nano-structures. In this dissertation, we study four closely related aspects of DNA based self-assembly and nano-devices: complexity of self-assembly, fault-tolerant self-assembly, DNA robotics devices, and DNA computing devices.

Complexity of self-assembly . We establish a framework that models assemblies resulting from the cooperative effects of repulsion and attraction forces in a general setting of graphs. By capturing a much wider range of interesting self-assembly phenomena, it advances previous work that models simple rectangular grid structures formed by only attraction force. We define an accretive graph assembly model and a self-destructible graph assembly model, and obtain several complexity results including the first PSPACE-complete result in the study of self-assembly.

Fault tolerant self-assembly . Fault tolerance is essential for building complex synthetic self-assembled systems at the molecular scale. In the practical context of algorithmic DNA tiling lattices, we propose an information encoding scheme using overlaid redundant computation, which, for the first time, reduces the error rate from e to e 3 without increasing the size of the assembled lattice.

DNA robotics devices . A major challenge in nanotechnology is to precisely transport a nanoscale object from one location on a nano-structure to another location along a designated path. To address this challenge, we design DNA motors capable of autonomous, unidirectional, progressive linear motion along self-assembled DNA tracks. The practicality of the designs is partially supported by the experimental construction of a three-anchorage autonomous DNA walking device.

DNA computing devices . Building on the designs of the above robotics devices, we obtain the designs of autonomous DNA mechanical computing devices embedded in DNA lattices. These devices represent a novel converging point for studies on nano-lattice assembly, nano-robotics, and nano-computing. In particular, we present the designs of an autonomous universal DNA Turing machine and an autonomous universal DNA cellular automaton.

Indexing (document details)


Reif, John H.


Duke University

School Location:

United States -- North Carolina


Self-assembly, Nanorobotics, Molecular computations


DAI-B 66/06, p. 3248, Dec 2005

Source type:



Computer science

Publication Number:

AAT 3181510