Robust Composite DNA Storage under Sampling Randomness, Substitution, and Insertion-Deletion Errors

This project represents a highly specialized intersection of information theory and synthetic biology. By treating 'composite DNA' (where multiple nucleotides occupy a single position in a mixture) as a multinomial channel and applying digital modulation techniques (similar to QAM in telecommunications), it provides a mathematical bridge for increasing DNA storage density. The defensibility is low (4) from a software perspective because it currently has zero community traction (0 stars) and functions as a reference implementation for an academic paper. However, the intellectual moat is significant; replicating this requires deep expertise in both coding theory and biochemistry. Frontier labs like OpenAI or Google are unlikely to enter this space directly, as it is heavily dependent on hardware (DNA synthesis and sequencing) which is currently dominated by players like Illumina, Twist Bioscience, and startups like Catalog DNA. The primary risk is not from AI labs, but from the hardware providers themselves integrating these coding schemes into their proprietary stacks. The displacement horizon is long (3+ years) because DNA storage remains in the research/pilot phase and has not yet reached commercial commodity status.