DNA Encoding Codecs
Binary-to-nucleotide mapping with configurable density, GC balancing, and homopolymer constraints. Multiple codec strategies for comparing information density versus recoverability.
Research program
Computational DNA Storage Research
DNA Vault is a research platform for exploring how digital information can be encoded into molecular symbolic form, stressed under simulated sequencing noise, and recovered with verifiable integrity.
All experiments are computationally simulated. Wet-lab validation not currently claimed.
Research overview
DNA information storage sits at the intersection of information theory, error-correcting codes, molecular biology constraints, and sequencing technology. DNA Vault provides a software laboratory to model these interactions before physical integration.
The research program focuses on reproducible experiments: every run produces a manifest with engine version, configuration parameters, error model settings, recovery metrics, and verification status.
Algorithms
Reed-Solomon ECC
Block-level error correction with configurable symbol counts. Corrects substitution and burst errors within encoded chunks.
Fountain Coding
LT fountain code assembly for recovery from partial read sets. Configurable overhead, block size, and degree distribution.
Sequencing Channel Model
Parameterized simulation of substitution, insertion, deletion, dropout, and coverage — with seeded randomness for reproducibility.
Sync Framing
Frame synchronization for assembling simulated reads into recoverable blocks under indel and dropout conditions.
Verification
SHA-256 hash comparison providing cryptographic proof of bit-identical recovery after the full pipeline.
Experiments
The experiment engine runs the complete workflow: input → encode → simulate → recover → verify. Each experiment receives a unique identifier and produces a downloadable manifest.
INPUT↓
ENCODE↓
SIMULATE↓
RECOVER↓
VERIFY
Technical notes
Information density
Bits-per-base varies by codec, ECC overhead, fountain redundancy, and GC constraints. The density explorer in Studio calculates estimates from actual engine configuration — labeled as theoretical/computational.
Error sensitivity
Recovery rate depends on the interaction between ECC strength, fountain overhead, and the sequencing error model. Systematic sweeps across error parameters reveal configuration boundaries.
Reproducibility
Seeded experiments produce identical results given the same input, configuration, and engine version. Manifests capture all parameters needed to reproduce a run.
Hosted limits
The hosted platform enforces file-size and fountain-code limits. Large jobs can be run locally via the CLI. Limits are documented, not hidden.
Open problems
Active research questions the platform is designed to help explore.
Optimal redundancy under realistic error models
How much fountain overhead is required for reliable recovery at various substitution, indel, and dropout rates?
Codec density vs. biological constraints
What is the tradeoff between information density and synthesis/sequencing compatibility when GC and homopolymer limits are enforced?
Bridging simulation to physical validation
Which computational error models best predict recovery behavior in real sequencing data? Physical validation remains a future research direction.
Large-scale archival architectures
How should molecular data vaults be structured for petabyte-scale archival — addressing, indexing, and partial retrieval?
Research log
Technology phases in the DNA Vault research program. Phases are ordered by development progression — no specific dates are assigned.
Concept
Computational DNA information storage — framing the problem of transforming digital data into molecular symbolic representation and back.
Encoding Engine
Digital data → DNA symbolic sequences with configurable codecs and chunking.
Error Correction
Reed-Solomon ECC and fountain coding for resilient recovery from corrupted reads.
Sequencing Simulator
Synthetic sequencing noise models — substitution, indel, dropout, and coverage.
Recovery Engine
Bit-identical reconstruction from simulated reads through ECC and fountain assembly.
Web Platform
Interactive DNA storage laboratory — DNA Vault Studio with experiment console and Observatory metrics.
Benchmarking
Comparative experiments across codec and configuration dimensions. Infrastructure in development.
Physical Validation
Future phase — integration with synthesis and sequencing workflows for wet-lab validation.
DNA Data Infrastructure
Long-term vision — molecular data vaults as part of the world's information infrastructure.