AeonScript sits at the intersection of biology, information theory, and software. This glossary makes that vocabulary accessible from either side.
A, C, G, T — the four nucleotide bases of DNA: adenine, cytosine, guanine, thymine. A pairs with T; G pairs with C.
Base — a single nucleotide. In AeonScript, the smallest unit of information on a strand.
Bioinformatics — computational analysis of biological data (DNA, RNA, protein sequences).
bp — base pair. The length unit for DNA: 1 bp = one base on each of two complementary strands.
CDS — coding sequence. A region of DNA that codes for a protein.
Codon — a group of 3 consecutive bases that codes for one amino acid in the genetic code. AeonScript does not use codons; we encode binary data, not amino acids.
Complementary strand (reverse complement) — the strand pairing with a given DNA strand. If forward reads ACGT, the reverse complement reads ACGT (because A↔T and C↔G, then reversed).
dNaM / d5SICS / dTPT3 — unnatural nucleotides synthesized for expanded genetic alphabets (e.g. Romesberg lab, Synthorx). Future L1 profiles may use these.
DNA storage — the practice of encoding arbitrary digital data onto synthetic DNA, then reading it back via sequencing. Distinct from biological gene expression.
GC content — percentage of bases in a sequence that are G or C. Synthesis and sequencing both require GC content in a specific range (typically 40–60%, AeonScript v0.1 relaxes to 25–75%).
Helix — the iconic double-stranded shape DNA naturally takes.
Homopolymer — a run of identical bases (e.g. AAAA). Long homopolymers (typically > 6–10) confuse sequencing technologies. AeonScript v0.1 limits runs to ≤ 8.
IGSC — International Gene Synthesis Consortium. A coalition of synthesis vendors with a harmonized screening protocol to detect dangerous sequences. Referenced in SPEC §9.
Methylation — chemical addition of a methyl group (-CH₃) to a base. 5-methylcytosine (5mC) is the most common; readable by Nanopore sequencing. Future L1 profile L1-5m will use it as a fifth symbol.
Nanopore — sequencing technology (Oxford Nanopore) that reads bases as they pass through a tiny pore. Can detect methylated bases natively.
Nucleotide — synonym for "base" in this context.
Oligo / oligonucleotide — a short synthesized DNA sequence (typically 100–300 bases). AeonScript blocks are stored as collections of oligos.
ORF — open reading frame. A region of DNA that could code for a protein. Not relevant to AeonScript directly but referenced in §5.2 of the spec (multiplexed reading frames).
PCR — polymerase chain reaction. A technique for selectively amplifying a specific DNA sequence. Used in DNA storage to retrieve specific oligos via primer-based addressing (L3).
Primer — a short DNA sequence (~20 bases) used to mark and selectively amplify a region of DNA. AeonScript L3 will use PCR primers as block addresses.
Synthesis — manufacturing DNA from scratch (chemical or enzymatic). The current bottleneck for DNA storage economics.
Sequencing — reading existing DNA. Mature technology (Illumina, Nanopore, PacBio).
5mC, 5hmC — 5-methylcytosine, 5-hydroxymethylcytosine. Modified cytosines used as extra symbols in extended L1 profiles.
Bit — fundamental unit of information (0 or 1). AeonScript encodes 2 bits per DNA base in the canonical L1-4 profile.
Channel — abstract medium that transmits information with some error probability. DNA is the channel here.
Codec — encoder + decoder pair. AeonScript v0.1 is one codec; we hope many implementations exist.
Codeword — the output of an error-correcting code: data + parity. RS(255, 223) means each codeword is 255 bytes (223 data + 32 parity).
Compression — removing redundancy from data to reduce size. Different from error-correction (which adds redundancy).
Erasure — a known-missing symbol. Easier to correct than unknown errors. RS can recover twice as many erasures as errors.
Error correction (FEC, forward error correction) — adding redundancy so the receiver can recover from corruption without re-transmission. Reed-Solomon, LDPC, fountain codes are common.
GF(256) — Galois Field of 256 elements. The arithmetic system Reed-Solomon operates over. Each element is a byte.
LDPC — Low-Density Parity-Check. A modern error-correcting code, used in 5G and Wi-Fi. AeonScript L4 v0.2 will use LDPC as outer code.
Pseudo-random binary sequence (PRBS) — a deterministic stream that looks random. Used by AeonScript as a scrambler to balance GC content without adding entropy.
Reed-Solomon (RS) — classic error-correcting code (Reed & Solomon 1960). Used in CD audio, QR codes, satellite links. RS(n, k) encodes k data bytes into n bytes with (n−k)/2 errors correctable. AeonScript uses RS(255, 223) → corrects up to 16 errors per 255-byte codeword.
Run-length-limited (RLL) — encodings that bound the maximum run of identical symbols. Common in magnetic storage. Future AeonScript L1 profiles will use RLL.
Shannon limit — the theoretical maximum information rate of a channel. For DNA, that's 2 bits/base in the simplest model. Real codecs achieve ~70–90% of this due to constraints.
Wire format — the on-the-wire (or "on-the-strand") byte/base representation of a structured message.
API — Application Programming Interface. The functions an implementation exposes to callers.
CC-BY-SA — Creative Commons Attribution-ShareAlike. The license used for AeonScript documentation. Derivative works must remain under the same license.
Conformance — formal property that an implementation correctly follows the spec. See CONFORMANCE.md.
FAQ — Frequently Asked Questions. See FAQ.md.
MIT — permissive software license. Used for AeonScript code. Allows commercial use, modification, redistribution.
Normative — describing requirements that MUST be followed by conformant implementations. RFC 2119 keywords (MUST, SHOULD, MAY) mark normative text.
OSI model — 7-layer reference architecture for networking. AeonScript borrows the layered structure but maps it to DNA storage.
RFC 2119 — IETF document defining "MUST", "SHOULD", "MAY" in standards.
Test vector — a canonical input/output pair that conformant implementations MUST reproduce.
Block — a unit of storage that is independently addressable. One block = one L5 tag + payload + parity.
L1, L2, ..., L7 — the seven layers of the AeonScript architecture. See SPEC.
L1 profile — variant of the physical layer. L1-4 uses 4 symbols (A,C,G,T), L1-5m uses 5 (adds 5mC), L1-6 uses 6.
Permutation prefix — a 4-base sequence at the start of each oligo that tells the decoder which {A,C,G,T}↔{00,01,10,11} mapping the rest of the oligo uses. AeonScript v0.1 tries up to 5 permutations to satisfy biochemical constraints.
Phylogenetic tag — (not used in AeonScript) — in genomic foundation models (like Evo 2), a metadata string identifying species. AeonScript's semantic tag serves a different purpose.
Scrambler — the xorshift32 PRBS that XORs the stream before L1 encoding, to balance GC content.
Semantic tag — the |AEONSCRIPT=0.1;L1=L1-4;TYPE=...| prefix at the start of each block. Makes the file self-describing.
Self-describing — every block carries its own codec version, so it can be decoded without external context.
Missing a term? Open an issue and we'll add it.