Quantitative Assessment of Randomized DNA Base Sequences Using Multi-Model Physical Analysis for High-Fidelity Data Storage.

DNA is emerging as a promising medium for ultra-dense, long-term digital data storage, yet sequence design remains hindered by homopolymer formation and compositional bias, which compromise synthesis, sequencing, and decoding accuracy. Here, the study introduces a quantitative framework to evaluate and optimize randomized DNA base sequence design rules using three physics-inspired models: translational and rotational active particle trajectories, the inverse Ising model, and a 3-input 1-output logic algorithm system. Encoding schemes with varying homopolymer constraints are systematically applied to binary image data. Rigorous analysis reveals that stringent randomization rules markedly reduce homopolymer length, balance GC content, and enhance sequence randomness. Experimental validation via polymerase chain reaction (PCR) amplification and Sanger sequencing confirms high decoding fidelity (95-98%). This multi-model assessment establishes a robust strategy for designing DNA sequences with superior stability, reliability, and scalability for future molecular data storage systems.