DNA-SaM, a robust system for large-scale data storage

Abstract

DNA data storage offers a viable strategy to address the impending data explosion. Early attempts to harness DNA as a storage medium have encountered scalability limitations, largely due to the complexity of codec algorithms, the generation of biochemically harmful sequences and lack of a robust architecture. We present "DNA-SaM", a novel system designed for DNA data storage, which achieves linear computational complexity and strict bio-constraint adherence, ensuring high coding efficiency and fidelity. It encoded data at speeds surpassing classic systems by over 2 orders of magnitude, with this superiority changes across various encoding algorithms. Importantly, DNA-SaM effectively eliminates any sequence that could be deleterious to in vitro and in vivo biochemical processes, including homopolymer runs, tandem repeat motifs, and potential promoter sequences, etc. It also involves an advanced DNA data storage architecture that incorporates a two-tiered indexing system and a novel "storage unit" distribution paradigm for large-scale data storage. It is further validated by practical data storage both in vitro and in vivo with a 100% success rate. Our system is capable of storing data over 1039 PB, which marks a critical advancement in the scalability of DNA-based data storage.