Improving reliability in DNA-based computations

DNA computing is an attractive alternative to traditional computing technologies due to DNA´s large storage capacity, massive parallelism and energy efficient computation. Massive parallelism, in particular, makes DNA computing attractive for cryptographic applications. But since DNA computing is based on bio-molecular reactions, it is error prone. A main source of errors is hybridizations between unintended sequences during the computation process. To minimize these errors we can design error-resistant DNA sequences. Here we have used simulated annealing (SA) to design such sequences. Our simulations suggest that sequences generated by our approach are superior to sequences generated by all techniques previously reported.