Skip to content

Library for combinatorially specifying, constraining, and exploring genetic design spaces.


Notifications You must be signed in to change notification settings


Repository files navigation


Build Status npm versionCoverage Status

Latest stable version available at ConstellationCAD



node python

Local UI

git clone
npm run build && npm run start

Then open http://localhost:8082/ on browser

NPM Package

The package can be installed in the following way.

npm install constellation-js

It is possible to generate a collection of designs that match a specification using graph construction and traversals.

const constellation = require('constellation-js');
let goldbar = '{PT7_a then galK}';
let categories = {
	  "promoter": [
	  "cds": [
let result = constellation.goldbar(goldbar, categories, {designName: 'my-first-design'});
// result.stateGraph, result.designs, result.sbol, result.messages
Optional parameters Description
designName Name of design space for SBOL output, defaults is "constellation-design"
numDesigns Max number of designs to enumerate, default is 20
maxCycles Cycle depth for -orMore operators, default is 0
representation Choose between EDGE or NODE based graph, default is EDGE
Output Description
stateGraph See Graph Data Structure
designs List of enumerated designs
sbol See Synthetic Biology Open Language
messages Errors or warnings, if applicable

It is also possible to generate a collection of designs that match a specification using a purely symbolic approach (note that this approach supports only a tolerance of 0 for the AND operator and does not support the MERGE operator).

let result = constellation.symbolic(
               "(one-or-more a) then (one-or-more x)", 
               {"a": {"b": ["c"]}, "x":{"y": ["z1", "z2", "z3"]}}, 
               {"numDesigns": 'all', "maxCycles":7});

Case Studies

GOLDBAR syntax for the case studies described in the manuscript are available here and can be demoed on Constellation's UI via the drop down menu.

Design Space Representations

Genetic design spaces in Constellation are represented in three ways:

  2. Directed cyclic graph
  3. SBOL


The supported GOLDBAR concrete syntax for genetic design spaces is presented below using extended BNF notation. Notice that then and . are equivalent, and the delimiters (...) and {...} are equivalent.

 <seq> ::= <exp> then <seq>
        |  <exp> . <seq>
        |  <exp>

 <exp> ::= <term> or <exp>
        |  <term> and0 <exp>
        |  <term> and1 <exp>
        |  <term> and2 <exp>
        |  <term> merge <exp>
        |  <term>

<term> ::= one-or-more <term>
        |  zero-or-more <term>
        |  zero-or-one <term>
        |  reverse-comp <term>
        |  ( <seq> )
        |  { <seq> }
        |  <atom>

<atom> ::= ([A-Za-z0-9]|-|_)+

The JSON schema for the GOLDBAR abstract syntax tree representation (parsed from the concrete syntax presented above) can be found in schemas/ast.schema.json.

Graph Data Structure

Constellation supports both NODE and EDGE based versions of a design space. Below are examples equivalent to the GOLDBAR specification promoter then one-or-more CDS then terminator

node-graph edge-graph
Visualization of node-based graph Visualization of edge-based graph

The JSON schema for a design space graph can be found in schemas/graph.schema.json. Below is an example of a node-based graph within a single node in JSON format.

  "604571a7-9e38-4647-afd0-9546399480b5": {
    "id": "604571a7-9e38-4647-afd0-9546399480b5",
    "text": "root",
    "type": "root",
    "edges": [
    "operator": ["One"]

Synthetic Biology Open Language

SBOL is an open standard for the representation of in silico biological designs, and the icons used in this tool are provided by SBOL Visual. Design spaces are expressed in SBOL via the CombinatorialDerivation extension and can be exported and stored in Knox. This third form of design space representation allows Constellation to be easily integrated in the synthetic biology community.