Uploaded August 2026 | Updated November 2026, 5 hours ago
Recent advances in DNA strand displacement (DSD) circuits have expanded the scale and complexity of molecular computation. Several approaches to near-data information processing in DNA data storage have also emerged, including content-based retrieval with hybridization, in-memory computation over topologically-encoded data with DSD, and Boolean search with fluorescence sorting. However, practical integration of computation and storage remains limited by our ability to further scale up DNA circuits and develop compatible interfaces. Here, we scale up DSD logic circuits to perform Boolean search over a sequence-encoded DNA database. Both data and logic gates are encoded in compatible formats and can be derived from a single array-synthesized oligonucleotide pool through PCR amplification followed by nicking enzyme digestion. We use nicked double-stranded DNA gates to perform computation and introduce a novel data storage format where each “data-gate” contains a file payload fragment and unique file address. Accessed data-gates are distinguished from stored data through toehold exchange of the address strand followed by irreversible addition of a universal end-cap domain. We experimentally perform random access through Boolean search in a small DNA database of 16 encoded Haiku files. Each file is split between five or six payloads across 69 distinct data-gates. We implement an address decoder circuit with 16 parallel AND gates that map each combination of two-attribute inputs (e.g. poet: Bashō AND season: Autumn) to a unique address output. This enables parallel evaluation of Boolean database queries, where file subsets are selected based on conjunctive attribute constraints. We execute all 226 valid queries and read out relative data-gate activation with next-generation sequencing. Even for the most complex queries, involving seven simultaneous attributes and activating 12 of 16 files (e.g. SELECT * FROM haiku WHERE season ≠ “Summer”), we observe clear separation between target and off-target data-gates for each file interval (Fig. 1). To explore the scalability of this approach, we additionally demonstrate a circuit of 1,056 parallel join gates mapping 33 inputs to 10 parallel fork gates. This construction is structurally analogous to the Boolean search system, suggesting that this approach could scale to conjunctive queries over 33 file attributes addressing up to 10 files. Together, these results demonstrate a scalable and modular framework unifying DSD computation and high-density data storage.
Presented by
Chandler Petersen, University of Washington
This is a presentation from the 2026 Storage and Computing with DNA Conference.
· Learn More about the SNIA DNA Data Storage Alliance: snia.org/groups/snia-dna-technology-affiliate
· SNIA Educational Library: snia.org/library
· X: twitter.com/SNIA
· LinkedIn: linkedin.com/company/snia
Recent advances in DNA strand displacement (DSD) circuits have expanded the scale and complexity of molecular computation. Several approaches to near-data information processing in DNA data storage have also emerged, including content-based retrieval with hybridization, in-memory computation over topologically-encoded data with DSD, and Boolean search with fluorescence sorting. However, practical integration of computation and storage remains limited by our ability to further scale up DNA circuits and develop compatible interfaces. Here, we scale up DSD logic circuits to perform Boolean search over a sequence-encoded DNA database. Both data and logic gates are encoded in compatible formats and can be derived from a single array-synthesized oligonucleotide pool through PCR amplification followed by nicking enzyme digestion. We use nicked double-stranded DNA gates to perform computation and introduce a novel data storage format where each “data-gate” contains a file payload fragment and unique file address. Accessed data-gates are distinguished from stored data through toehold exchange of the address strand followed by irreversible addition of a universal end-cap domain. We experimentally perform random access through Boolean search in a small DNA database of 16 encoded Haiku files. Each file is split between five or six payloads across 69 distinct data-gates. We implement an address decoder circuit with 16 parallel AND gates that map each combination of two-attribute inputs (e.g. poet: Bashō AND season: Autumn) to a unique address output. This enables parallel evaluation of Boolean database queries, where file subsets are selected based on conjunctive attribute constraints. We execute all 226 valid queries and read out relative data-gate activation with next-generation sequencing. Even for the most complex queries, involving seven simultaneous attributes and activating 12 of 16 files (e.g. SELECT * FROM haiku WHERE season ≠ “Summer”), we observe clear separation between target and off-target data-gates for each file interval (Fig. 1). To explore the scalability of this approach, we additionally demonstrate a circuit of 1,056 parallel join gates mapping 33 inputs to 10 parallel fork gates. This construction is structurally analogous to the Boolean search system, suggesting that this approach could scale to conjunctive queries over 33 file attributes addressing up to 10 files. Together, these results demonstrate a scalable and modular framework unifying DSD computation and high-density data storage.
Presented by
Chandler Petersen, University of Washington
This is a presentation from the 2026 Storage and Computing with DNA Conference.
· Learn More about the SNIA DNA Data Storage Alliance: snia.org/groups/snia-dna-technology-affiliate
· SNIA Educational Library: snia.org/library
· X: twitter.com/SNIA
· LinkedIn: linkedin.com/company/snia










