Cyclic Mixture Mutagenesis
for DNA-Based Computing

Abstract:

Leonard Adleman recently demonstrated that chemical reactions involving DNA could solve a limited class of computational problems [Adleman-94]. The research program described in this document was inspired by the goal of developing a better DNA computation system, one that would be capable of executing arbitrary computations, while taking advantage of astronomical numbers of molecular, DNA-based nanoprocessors all working in parallel.

In the work performed so far, a design for a new DNA-based information processing system called cyclic mixture mutagenesis, or CMM, has been worked out and applied successfully in computer simulations based on empirically-derived models of DNA interactions from the biochemistry literature. CMM is designed to be computation-universal in theory, and to be easy to execute in practice at any molecular biology laboratory. A series of lab experiments designed to test the DNA chemistry models and refine the relevant laboratory protocols is already in progress.

The two primary goals of the dissertation research program proposed here are, first, to continue the lab work in an attempt to prove that CMM is capable of carrying out certain simple computations that are of interest to biologists and others that are of interest to computer scientists, and second, to further develop the theory behind the technique and explore variations designed to yield improvements in the method's biological feasibility or computational power. Subsidiary goals of the dissertation research include considering potential applications of CMM in biology and medicine, conducting a pcomparative survey of other molecular computation techniques, and analyzing the possible implications of DNA computers for the long-term progress of higher-speed and smaller-scale computing.

Contents: