Molecular Computation With Automated Microfluidic Sensors (MCAMS)

We combined microfluidic technology with recently developed algorithms of nucleic acid-based computing to demonstrate a compact, automated nucleotide-based computational device capable of rapid detection of computational output. We explored both experimental and theoretical aspects of the construction of microfluidic nucleic acid computing. Microreactors implementing Boolean operators lent themselves to a relatively simple implementation of DNA computing. On the theoretical side, we simulated the performance of individual reactors and designed small reactor systems (sorters) mimicking AND and OR gates and systems capable of solving complex Boolean expressions. We compared positive vs. negative selection and parallel vs. series arrangements of reactors, considering both the accuracy of and disproportionate representation ( bias ) within the final pool of solutions. We also developed O(P2) software that designs operating instructions for an O(NL) system capable of solving any Boolean expression. We developed software for automated data collection and signal analysis of electronic detection of labeled DNA and explored novel nanocrystal probes that enable detection of DNA for readout in on-chip electrical assays. We focused in the areas of methods to produce nanocrystals bearing precise numbers of oligonucleotides, and development of nanocrystal precise groupings that bear a collective signature.
The original document contains color images. Funded in part by DARPA.