Team:UCSD Software/Project

Abstract

Genetic circuits are often difficult to engineer, requiring months to design, build, and test each individual genetic device involved in the circuit. SBiDer, a web tool developed by the UCSD Software iGEM team, will leverage existing devices to construct a database with consideration for the function of each device interpreted as boolean logic. The data can be queried by the user through SBiDer's visual interface to explore circuit designs. The displayed circuit's literature reference, characterization data, and images of included devices can be viewed through the built-in table. Basic validation of the circuit performance is also provided within in the interface. SBiDer's web of information can be expanded through user-generated additions to the database to improve the efficiency of the application and the accuracy of the models. (Partial STUB PAGE)

Future Directions

Synthetic genetic circuits created by synthetic biologists have yielded exciting applications such as biofuels production and cancer killing bacteria. These circuits are often difficult to engineer, requiring months to design, build, and test each individual genetic device involved in the circuit. Although there are many genetic devices that have been built, re-using these devices often requires a time-consuming review of the literature. The UCSD Software iGEM team will address this challenge by creating a web-tool that leverages existing genetic devices to create complex genetic circuits. (STUB PAGE)

Web Application

We will connect known genetic devices together via device input and outputs to create a network of devices that can interact. We define a genetic device as a DNA construct transformed into cells that can cause expression of some protein in response to stimuli (or input). We will develop a web interface to facilitate access to the complex network that we have constructed. Our Web interface makes extensive use of Cytoscape, an open source bioinformatics software package for metabolic network visualization and simulation. In addition, the interface will generate SBOL Visual Images, a standard language that is easily understood by synthetic biologists all over the world. Users can also update our database with additional devices through this interface. Using the Cynetshare framework, users can share their circuit designs. (stub page - going to write about the technologies that we used and why we used them).

The Search Algorithm

Introduction to the SBiDer Search Algorithm

Purpose of the Algorithm

SBiDer is fundamentally constructed on a manually curated database of existing genetic devices, or operons. This SBiDer database stores operons, species, and most importantly the biochemical interactions of these elements. Also, the database can be represented as an operon interaction network via species. Using this network representation of the database, SBiDer’s search algorithm searches for feasible operon paths connecting a set of species to another set of species. As a result, given a set of input species and a set of output species, the search algorithm returns a subnetwork of operon paths connecting the input species to the output species. Each path within the subnetwork is an operon path that can be used to produce the output species from the input species. In other words, placing operons from a returned operon path as well as the input species into a system can produce the output species.

Biology Captured by the Algorithm

The search algorithm conducts a complex, effective, and robust operon path search over the network representation of the database, ultimately generating single or multiple operon paths connecting a set of input species to a set of output species. A resulting path represents a system that can produce the output species given the input species. Also, each path attempts to capture biological reality by considering two major mechanisms of operon activation (detailed description below). Finally, SBiDer represents a resulting path using Petri net model, such as chemical reactions (C.A. Petri and W. Reisig). The Petri net generated by the search algorithm consists of all of the input species, input transitions, input transition logic, operons, output transitions, and output species in a path. As a result, the Petri net model captures the fundamental components of operon interactions, allowing the biological phenomena to be better understood, modeled, and analyzed.

Robustness and Modularity of the Algorithm

Due to the current number of operons in the database, the search algorithm should scale with the increasing size of the database, which will expand as the synthetic biology community adds more species, operons, and plasmids to the database (see complexity analysis of the algorithm is located below). In addition to this robustness, the search algorithm is easily modifiable since it has been implemented independently from the database and the SBiDer web application. Therefore, the search algorithm can be applied to a broad range of networks. Furthermore, the search algorithm was developed on an open source platform to encourage global collaboration in further optimizing the search algorithm and minimizing the barrier for algorithm improvement. As a result, the database as well as the search algorithm will be easily, effectively, and accurately expanded, extended, and optimized by the global synthetic biology community - together.

Database (the foundation)

We will mine the scientific literature for existing genetic devices and then construct a database that captures device characteristics such as:

• composition of devices
• function
• characterization data
• literature reference

We will design our database by rigorously constructing an entity relationship diagram and then normalizing these relationships to construct tables for a relational database. (another stub)

Modeling

We will connect known genetic devices together via device input and outputs to create a network of devices that can interact. We define a genetic device as a DNA construct transformed into cells that can cause expression of some protein in response to stimuli (or input). We will develop a web interface to facilitate access to the complex network that we have constructed. Our Web interface makes extensive use of Cytoscape, an open source bioinformatics software package for metabolic network visualization and simulation. In addition, the interface will generate SBOL Visual Images, a standard language that is easily understood by synthetic biologists all over the world. Users can also update our database with additional devices through this interface. Using the Cynetshare framework, users can share their circuit designs. (stub page - same as web app stub page).