Library based motif compilation of biological functionalityView all posters
Boston University, United States
One of the major design challenges of Synthetic Biology has been in designing novel genetic devices. Although many such devices have been currently built and tested in-vivo, a major issue faced is in translating a desired behavior to a combination of available genetic Devices quickly, accurately and automatically. Tools that can accomplish this will free designers to focus on the assembly and test of larger systems. In particular, these tools should propose multiple potential implementations with different DNA parts, technologies, and performance driven by user defined “motif libraries”. We present a software workflow called “Cello” to encapsulate Devices as independent modules and combine them to produce biological systems satisfying a more complex behavior. First, we describe “Cello-Motif”; a method for specifying Device composition, its intended function and rules for device to Device combination. Next, we present an algorithm to automatically generate candidate biological circuits satisfying a given specification using a library of Cello-Motifs. The specification can be either full or partial specified Boolean logic. The algorithm utilizes user-defined constraints: Upper-bound on the length of the circuit and the number of Parts of a particular type to be used – to produce an optimal set of candidate circuits. The components of the resultant candidate circuits are assigned Parts from a database, and the dynamic behavior of the circuit can be simulated if biophysical models of the underlying Parts are available. We provide transcriptional, translational and recombinase based “Cello-Motif” examples and the resultant candidate circuits generated for a given specification. This demonstrates that the approach is sufficiently powerful to describe functionality and composition of Devices, while independent of the underlying biological mechanism realizing the function. We also demonstrate methods to use simulated results to computationally predict the “best-match” circuit from the candidate set by asserting which result statistically best-fits the proposed specification.