Performance of DNA data embedding algorithms under substitution mutations

DNA data embedding is a relatively recent area which aims at embedding arbitrary information in deoxyribonucleic acid (DNA) strands. One interesting application of DNA data embedding can be tracing pathways of genetic material in novel ways. This paper explores the decoding performance of several DNA data embedding algorithms proposed in the literature, which are also briefly reviewed. DNA may undergo random mutations, which can cause errors at the decoding stage of such algorithms. Although some proposed methods do account for such errors, decoding performance under mutations has not been previously studied in general. The empirical performance comparison that we provide here allows to fairly compare a number of DNA data embedding algorithms under mutations for the first time. The evaluation is undertaken in silico by means of Monte Carlo simulations. Additionally, we propose two new DNA data embedding algorithms with good robustness properties.