Chapter 2
A synthetic biology registry has to answer a practical question: can a researcher who has never met the author of a part find it, understand what it does, build with it safely, and trust the result? Sequences are only the starting point. A promoter sequence with no record of its host, strength, safety level, or history is little more than a string of letters. WIA-BIO-021 defines the records that surround a sequence so that these questions have machine-readable answers.
Biological parts are reused far more often than they are rewritten. A team building a fluorescent reporter does not redesign a ribosome binding site from scratch. It looks for an existing one with known behavior. That only works if the registry offers four things:
The specification states its purpose as a unified framework for registration, storage, and sharing that preserves data integrity and reproducibility. It covers BioBrick and non-BioBrick classification, a metadata schema, sequence storage with format conversion, characterization structures, version control, access control and licensing, and integration with iGEM, Addgene, and SynBioHub. The terms it fixes are used throughout the rest of this chapter. A part is a discrete DNA or RNA sequence with a defined function. A device combines several parts into an integrated function. A BioBrick is a standardized genetic part assembled under RFC10 or RFC25. Characterization is experimental data describing how a part behaves.
The classification system divides parts into seven primary types. Each type carries its own set of expected properties, so a registry can ask the right questions of each record. A promoter is described by strength, not by molecular weight, and a coding sequence is the reverse.
| Type code | Category | Typical properties | Spec example |
|---|---|---|---|
| PRO | Promoter | Strength in relative fluorescence units, constitutive or inducible, organism specificity, regulatory mechanism | BBa_J23100, BBa_R0011 (LacI-regulated) |
| RBS | Ribosome binding site | Translation initiation rate, organism compatibility, spacing requirements, secondary structure | BBa_B0034 (medium), BBa_B0030 (strong) |
| CDS | Coding sequence | Protein product, molecular weight, activity, expression conditions, codon optimization status | BBa_E0040 (GFP), BBa_E1010 (mRFP1) |
| TER | Terminator | Termination efficiency, directionality, context dependency | BBa_B0015 (double terminator), BBa_B0010 |
| VEC | Plasmid backbone | Antibiotic resistance, origin of replication, copy number, cloning sites, size in bp | BBa_pSB1C3, BBa_pSB1A3 |
| COM | Composite part | Component list, assembly method, expected function, validation status | BBa_K123456 (GFP expression device) |
| REG | Regulatory element | Binding protein or molecule, regulatory mechanism, affinity or binding constant | Operators, protein binding sites |
Backbones such as pSB1C3 and pSB1A3 are recorded as parts in their own right. This reflects how experiments are built: the resistance marker and copy number of the vector determine how a part is maintained, and they affect how its behavior should be interpreted.
Identifiers follow the pattern BBa_[Type][Number], for example BBa_K123456, a composite part from a team using the K series. The prefix letters look like a type indicator, but the material shows they are not a reliable one. The same letter is allowed for more than one type:
BBa_B is listed for both ribosome binding sites (BBa_B0034) and terminators (BBa_B0015, BBa_B0010).BBa_P or BBa_J, and coding sequences may use BBa_C or BBa_E.BBa_R prefix with ribosome binding sites.BBa_pSB[number] in the specification, while a shorter summary list abbreviates them as BBa_V*.This is why the standard stores the type as an explicit metadata field instead of deriving it from the identifier. Software that guesses a part's function from its first letter will misclassify real parts. A registry client should read type, and treat the identifier as an opaque, permanent label.
Parts do not appear fully formed. They are designed, synthesized, tested, and sometimes withdrawn. The standard records this progression as a status with a two-letter code.
| Status | Code | Meaning |
|---|---|---|
| Planning | PL | Part design phase |
| Ordered | OR | DNA synthesis ordered |
| Available | AV | Part available in registry |
| Characterized | CH | Experimental data available |
| Validated | VA | Independently verified |
| Deprecated | DP | Superseded or problematic |
| Deleted | DL | Removed from registry |
The distinction between Characterized and Validated matters scientifically. Data produced by the author is one thing, and verification by someone else is another. Deprecated keeps a record visible so that downstream users learn a part was superseded or found problematic, whereas Deleted removes it from the registry.
The sector has no single repository. The standard names three registries it interoperates with, and each plays a different role. iGEM hosts the BioBrick part collection tied to its competition. Addgene is known for distributing plasmids. SynBioHub is a repository built around SBOL-encoded designs. A researcher's data is therefore split across systems that use different identifiers, file formats, and levels of annotation.
The standard does not try to replace these services. It defines a common record that can be exported to and imported from them. The CLI shows this in practice: export --format sbol writes a part to an SBOL file, and import --format genbank reads a GenBank plasmid record. Interoperability here means that a record can leave one system and arrive in another without losing its type, sequence features, safety level, or license.
Applied work such as reporter devices, inducible protein production, and constitutive expression cassettes depends on knowing how parts perform. The standard's own composite examples are of this kind: BBa_K123456 is a GFP expression device and BBa_K654321 is an inducible protein production system. Choosing a promoter for such a device means comparing strengths, and promoter strength is expressed in relative fluorescence units. The registry therefore treats characterization as first-class data.
The CLI records it with a command such as wia-bio-021 characterize --id BBa_K123456 --strength 1500 --growth-rate 0.8. Searches can then filter on it, as in strength: 'high'. Strength and growth rate are only meaningful with their biological context, which the standard captures separately (see section 2.8). A strength value without a host, medium, and temperature cannot be compared across laboratories.
Every part carries a biosafety level, because open sharing makes it easy to circulate material that should not circulate freely. The specification defines four levels:
Because the level is a required field, a registry can withhold or gate high-risk parts instead of discovering the problem after distribution. Licensing is handled the same way. A license value such as CC-BY-4.0 travels with the record. For formats, parts are stored in the Synthetic Biology Open Language (SBOL), and the metadata is described as compatible with both SBOL and GenBank. A GenBank accession can be attached as an external link, so a registry record can point to the public sequence database entry it relates to.
A registered part must carry a required set of fields. The specification's example is condensed below.
{
"partId": "BBa_K123456",
"partName": "Strong GFP Expression Device",
"type": "composite",
"subtype": "reporter",
"status": "available",
"safetyLevel": "BSL-1",
"author": { "name": "Team iGEM 2024", "affiliation": "MIT" },
"created": "2024-06-15T10:30:00Z",
"modified": "2024-08-20T14:45:00Z",
"version": "1.2.0",
"license": "CC-BY-4.0"
}The sequence object sits beside this record. It holds the nucleotides, a length, a format (raw, fasta, genbank, or sbol), a sequenceType (dna, rna, or protein), and a checksum written as sha256:.... The checksum lets any consumer verify that the sequence it received is the one that was registered. A features array annotates regions with type, start, end, strand, and name. In the example, a J23100 promoter occupies positions 1 to 35 and a B0034 RBS positions 40 to 52, both on the forward strand. This is how a composite part records its own internal structure.
A third block holds the biological context: organism (Escherichia coli K12), strain (DH5α), and growth conditions including temperature, medium, antibiotics, and inducers. It also holds expression details such as host, cellular compartment, tags like His6, and fusion partners. Documentation fields, including design notes, DOI-based references, and external database links, complete the record.
The version-control mechanism points to where registries are heading. A part's version and its created and modified timestamps already allow a part's evolution to be traced. Combined with the Deprecated status, this lets a community retire a flawed part without erasing the history of experiments that used it. As more sites exchange SBOL and GenBank records, shared identifiers and well-formed checksums will matter more than any single repository.
Registry data connects three groups of people. Authors deposit parts and keep ownership under a chosen license. Builders search, retrieve, and assemble them. Reviewers and safety officers rely on the status and safety fields to decide what is acceptable. The same record serves all three, which is why the standard puts identity, evidence, and risk information in one structure instead of splitting them across tools.
The standard defines a BioBrick as a part that follows an assembly standard, naming RFC10 and RFC25. The RFC label stands for Request for Comments, the community document format in which BioBrick assembly rules are published. In the widely used RFC10 scheme, each part is flanked by a prefix and a suffix containing the EcoRI, XbaI, SpeI, and PstI restriction sites. Two parts can be joined because XbaI and SpeI leave compatible ends. Ligating them destroys both sites and leaves a short scar, while the outer EcoRI and PstI sites remain intact so the new composite is itself a BioBrick that can be assembled again. RFC25 is an alternative scheme that adds sites to allow in-frame protein fusions.
For registry design, the lesson is that the assembly method is metadata. Composite parts record a component list and an assembly method. A user who needs a particular scar behavior, or who wants to fuse proteins, must be able to filter on that information, and the record has to say which standard a part was built under.
iGEM is a student competition in synthetic biology, and one of its lasting outputs is a shared collection of parts contributed by participating teams. The spec's examples reflect this origin: the author entry Team iGEM 2024, and the composite series BBa_K, described as parts from a team. Each year's teams add composites built from earlier parts, and a registry's quality depends on how well those contributions are documented.
The weakness this creates is uneven documentation, since many contributors are short-term participants. The standard answers it by making the metadata fields mandatory, separating Available from Characterized and Validated, and preserving author, affiliation, and version history. A part that was merely deposited is not mistaken for one that has been verified.
The TypeScript SDK exposes searchParts and getPartDetails. The finder below searches for high-strength E. coli promoters and prints each one's recorded characterization. It assumes each search hit exposes a partId; the other calls are taken from the SDK examples.
import { searchParts, getPartDetails } from '@wia/bio-021';
async function findStrongPromoters() {
const results = await searchParts({
type: 'promoter',
organism: 'E. coli',
strength: 'high'
});
for (const hit of results) {
const details = await getPartDetails(hit.partId);
console.log(hit.partId, details.characterization.strength);
}
}
findStrongPromoters();The same task works from the shell: wia-bio-021 search --type promoter --organism "E. coli" lists candidates, and wia-bio-021 get BBa_K123456 retrieves one record. Before building, a careful user would also check safetyLevel and status on each record, and would skip anything marked deprecated.
The part types map onto familiar engineering goals. Reporter devices pair a promoter and RBS with a fluorescent protein such as GFP or mRFP1 to make gene expression visible. Inducible systems use regulated promoters, such as the LacI-regulated BBa_R0011, to control when a protein is made. Each of these designs reduces to a choice of parts and a choice of backbone, and the registry is the place where those choices are documented.
type as a field instead of relying on the identifier prefix? Give two examples of prefixes that would mislead a program.The challenge this chapter set out was to make shared parts reusable. The standard's answer has five elements:
The chapter also covered how RFC10 and RFC25 assembly, iGEM team contributions, and the SDK and CLI fit into that picture. With the vocabulary and the records now defined, the next chapter turns to the sequence itself: how DNA, RNA, and protein sequences are stored, how FASTA, GenBank, and SBOL representations are converted into one another, and how features and checksums keep a sequence trustworthy when it moves between registries.