Hybridization and ring optimization for larger sets of embeddable biomarkers

Abstract

Embeddable biomarkers are short strands of DNA that can be incorporated into genetic constructs to enable later identification. They are drawn from error correcting codes on the DNA alphabet relative to the Levenshtein metric. This study uses three types of evolutionary algorithms to improve the best known size of DNA error correcting codes, improving the bound for nine different code parameters. One of the algorithms is used on only one set of code parameters, correcting an oversight in an earlier study. The other two algorithms are a ring optimizer and a hybridizing evolutionary algorithm that exploits previously known codes. The ring optimizer improves two code size bounds and sets the stage for the hybridizer to improve four more. The hybridizer requires the results of a previous search as a starting point. Starting with known codes from earlier work, it improves a total of six bounds. The best results found by this algorithm used the results of the ring optimizer as a starting point. The paper discusses the issue of building a suite of cooperative code-search algorithms as a good target for future work.