Efficient and low-complexity variable-to-variable length coding for DNA storage.

Area of Science:

• Bioinformatics
• Data Storage
• Molecular Engineering

Background:

• DNA-based storage offers high capacity and longevity for growing data needs.
• Encoding data into DNA is constrained by homopolymer (consecutive identical bases) and GC content limits.
• Violating these constraints increases DNA sequencing and synthesis errors.

Purpose of the Study:

• To address the source coding problem in DNA data storage under homopolymer and GC content constraints.
• To develop a novel coding technique that adheres to these DNA sequence limitations.
• To achieve near-optimal data storage rates with efficient computational complexity.

Main Methods:

• Developed a novel variable-to-variable-length encoding method for DNA data storage.
• Ensured adherence to both homopolymer and GC content constraints.
• Maintained linear complexity for encoding, enabling scalability with increased block lengths.

Main Results:

• The proposed coding technique effectively handles homopolymer and GC content constraints.
• Achieved near-optimal data storage rates, demonstrated by reaching 1.988 bits/base (theoretical limit 1.990) for specific parameters.
• Validated the method's effectiveness on both random and existing data files.
• The associated code is publicly available on GitHub.

Conclusions:

• A novel variable-to-variable-length encoding method for DNA data storage has been developed.
• The method achieves near-optimal storage rates without relying on concatenating predefined sequences.
• This approach offers a more efficient and robust solution for DNA data archiving.