Methods of Evaluating Gene Expression Levels

Technology #15194

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Image Gallery
The TASBE workflow has four general stages: specification, compilation, part assignment, and assembly. These are further refined into seven models: (1) high-level program specifying the desired behavior; (2) dataflow graph representing a computation as a set of functional operators that produce and consume data values; (3) abstract genetic regulatory network (AGRN) consisting of a set of partially specified functional units; (4) genetic regulatory network (GRN), consisting of the same elements as an AGRN, except that all of the elements must be fully specified; (5) part sequence(s), where each sequence is a concatenation of available DNA parts with corresponding sample information; (6) assembly plan including a sequence of protocols for creating each part sequence from available DNA samples; (7) physical samples comprising DNA that actually instantiates a biological network. The tools transform each model in turn to its successor.Stages of compilation and optimization for mammalian (a,c,e,g) and E. coli (b,d,f,h) cellular platforms. For dataflow graphs (a,b,c,d), operators are shown as boxes, variables as arrows connecting from output to inputs and annotated with their data type, and function definitions as dotted boxes with the name of the function in the corner and a star on the function’s output variable. For AGRNs (e,f,g,h), functional units are low horizontal lines connecting promoters (bent arrows) and open reading frames (boxes) with terminators not shown on the diagram; activation and repression are green arrows and red stub connectors respectively, and all are annotated with the data types or values they represent.For each cellular platform, MatchMaker uses a feature database and signal level information (not shown) to select parts for instantiating the AGRN of the sensor/actuator test program. Parts selected by MatchMaker are marked in yellow.Plan generated by Puppeteer for BioBrick assembly of E. coli sensor/actuator test program (root at bottom) from available part samples (leaves at top). Edges represent restriction digests; dots represent ligations. The red dot ligation was executed robotically, while black dot ligations were executed by hand.
Professor Ron Weiss
Department of Biological Engineering, MIT
External Link (
Jacob Beal
Department of Biological Engineering
Aaron Adler
Department of Electrical Engineering and Computer Science, MIT
Noah Davidsohn
Department of Biological Engineering, MIT
Fusun Yaman-Sirin
BBN Technologies Corp.
Managed By
Jon Gilbert
MIT Technology Licensing Officer
Patent Protection

Methods of Evaluating Gene Expression Levels

US Patent 8,809,057
An End-to-End Workflow for Engineering of Biological Networks from High-Level Specifications
ACS Synthetic Biology, July 10, 2012, pp. 317-331


The ability to build a higher level representation of a novel biological design from known parts, with flexible protocol automation and DNA expression characterization, is useful in any industrial or academic setting where accurate predictions of the protein expression output of a genetic circuit is relevant to the project design.

Problem Addressed

The ability to accurately predict the expression levels of biological circuits is built on an understanding of the behavior of the individual elements involved in the biological circuit. Previous prediction methods relied heavily on human experience to supplement insufficient data. This design process is often inconsistent and difficult to replicate. A new system of tools that uses modular data on the input and response dynamics of individual circuit elements to accurately and consistently predict circuit behavior would greatly advance the field of knowledge and provide researchers with a systematic approach to the design and implementation of circuits. 


This technology includes a pipeline to characterize the expression levels of test proteins whose expressions are regulated by test regulatory elements. These test proteins are actively transcribed when a constitutively expressed effector-regulated protein is activated by an effector molecule and binds to the test regulatory element controlling the expression of the test proteins. Input reporter proteins controlled by the same test regulatory elements as the genetic elements are used to quantify the level of effector protein activity, while output reporter proteins controlled by the proteins encoded by the genetic elements are used to quantify the level of expression of the genetic element itself. The process for collecting data on this circuit includes first transfecting a population of cells with plasmids that contain separate elements of the genetic circuit, followed by the transfecting cells with a plasmid containing all the elements of the genetic circuit, then by integrating the entire circuit within the cells' chromosomes, and finally by transfecting cells with a non-replicable single-copy plasmid containing all the elements of the genetic circuit. At each step of the way, the cells are screened by fluorescence activated cell sorting (FACS) to measure the expression levels of input and output reporter proteins in order to determine the activities of the effector-regulated protein and test protein. The data, which can be collected as early as two weeks into the characterization process, are parsed and analyzed by the BioCompiler software system to determine the expression curve of the test proteins. A separate MatchMaker software system can use this data to predict the signal compatibility of test proteins when building a circuit containing multiple test protein elements, and the Puppeteer software system then automatically generates a protocol for a liquid-handling robot to implement the circuit designs in the laboratory. 


  • High-level representation of biological designs with BioCompiler
  • Automated DNA part assignment with MatchMaker
  • Flexible protocol automation with Puppeteer
  • High-throughput DNA element characterization method with multiple data collection points and fast turnover time