Shanghai Jiao Tong University researchers built a DNA storage codec named Siyuan Code that reaches theoretically optimal density, according to a peer-reviewed study published in Nature Communications. The experimental system demonstrated a storage density of 1.66 bits per nucleotide and a capacity of 41 exabytes per gram of DNA.
The codec uses a bijective coding method, creating a one-to-one mapping between binary data and low-error nucleotide sequences. Existing coding approaches often struggle to filter out error-prone sequences without discarding usable ones, which lowers density and requires higher strand copy counts. To resolve this trade-off, the team used a data structure to catalog and avoid error-prone patterns. This structure allows the system to operate under strict biochemical constraints with minimal redundancy.
Laboratory experiments showed that the system achieved perfect data retrieval using only six copies per strand. On commodity hardware, encoding and decoding speeds reached approximately 1 megabyte per second. The authors noted that this performance makes the processing throughput comparable to early USB flash drives.
Jiasen Li, Mingkai Dong, Zhengwei Qi, and Haibing Guan conducted the work across the School of Computer Science and the School of Electronic Information and Electrical Engineering at Shanghai Jiao Tong University. The National Key R&D Program of China and the Shanghai Key Laboratory of Scalable Computing and Systems funded the project, alongside support from the SJTU Kunpeng&Ascend Center of Excellence. The team also acknowledged Fei Wang for discussions covering DNA synthesis and sequencing fundamentals. Nature Communications received the paper on 27 June 2025 and published it on 27 August 2026 following acceptance on 13 August.
