A Robust and Efficient DNA Storage Architecture Based on Modulation Encoding and Decoding
Xiangzhen Zan,Ranze Xie,Xiangyu Yao,Peng Xu,Wenbin Liu
DOI: https://doi.org/10.1021/acs.jcim.3c00629
IF: 6.162
2023-06-08
Journal of Chemical Information and Modeling
Abstract:Synthetic DNA has been widely considered an attractive medium for digital data storage. However, the random insertion-deletion-substitution (IDS) errors in the sequenced reads still remain a critical challenge to reliable data recovery. Motivated by the modulation technique in the communication field, we propose a new DNA storage architecture to solve this problem. The main idea is that all binary data are modulated into DNA sequences with the same AT/GC patterns, which facilitate the detection of indels in noisy reads. The modulation signal could not only satisfy the encoding constraints but also serve as prior information to detect the potential positions of errors. Experiments on simulation and real data sets demonstrate that modulation encoding provides a simple way to comply with biological constraints for sequence encoding (i.e., balanced GC content and avoiding homopolymers). Furthermore, modulation decoding is highly efficient and extremely robust, which can correct up to ∼40% of errors. In addition, it is robust to imperfect clustering reconstruction, which is very common in practice. Although our method has a relatively low logical density of 1.0 bits/nt, its high robustness may provide a wide space for developing low-cost synthetic technologies. We believe this new architecture may boost the early coming of large-scale DNA storage applications in the future.
chemistry, multidisciplinary, medicinal,computer science, interdisciplinary applications, information systems