DNA-Based Indexing

Abstract

DNA has been acknowledged as a suitable medium for massively parallel computing and as a “smart” glue for self-assembly. In this paper, a third capability of DNA is described in detail as memory capable of encoding and processing large amounts of data so that information can be retrieved associatively based on content. The technique is based on a novel representation of data on DNA that can shed information on the way DNA-, RNA- and other biomolecules encode information, which may be potentially important in applications to fields like bioinformatics and genetics, and natural language processing. Analyses are also provided of the sensitivity, robustness, and bounds on the theoretical capacity of the memories. Finally, the potential use of the memories are illustrated with two applications, one in genomic analysis for identification and classification, another in information retrieval from text data in abiotic form.