interim

December 15th, 2006

Problem Definition:	Boolean networks are used to model the regulation of genetic network. Probabilistic Boolean networks, although based around probabilistic statistics, can provide accurate results in predicting the most efficient method of controlling dysfunctional networks. The purpose of this program is to create a computational tool that easily identifies targets of diseases, most importantly cancer, and provide a therapeutic intervention of the gene networks. Random and actual real life gene perturbations can be used to experiment the effectiveness of recognition and intervention processes.

Solution Process:	In order to accomplish a model structure and determine effective intervention methods, the problem sectionalized into three major procedures: Simulation, Inference, and Intervention. Structures that are defined are the Boolean network, the Boolean node (gene), and user specifications.

The purpose of the Simulation algorithm is to produce random, yet reliable arrays of a gene perturbation. These arrays are translated into Boolean terms (0’s and 1’s) within a truth table in order to create definite trajectories. Specifically, each node can be determined by the function of its parent(s).

The Inference procedure is an inverse function of the Simulation algorithm. There are two major steps in order to translate a trajectory of a Boolean network into truth tables in order to verify a node’s parent(s). The program applies the trajectories created by the Simulation algorithm and then applies the values to create a transition table. The transition table is an enumeration of the 0’s and 1’s. In order for a set of nodes to be valid parent(s) of a node, the enumeration of Boolean values of the affected node must coincide with a consistent value for each combination of the set of parent(s).

Most importantly to environmental situations is the Intervention algorithm. It is during this process that the identified perturbation in the Boolean network is altered to deem the network more ordinary. The program finds the most efficient way of determining the least amount of genetic changes in order to return the network to a state of normality. Simple as it may sound, sometimes, many gene modifications must be made which can prove to be costly and very time enduring.

Since the possible states of a single node is 2n, efficiency in determining a favorable gene therapy procedure is crucial. Thus, access to supercomputing resources will be the most resourceful way to advance.

Progress:	Up to this point, the data structure and file formats of the Boolean network, Boolean node, and specifications of the user have been determined. The structure and formats for these objects are the foundation to the entire program.

The Boolean network structure: The Boolean Network class has two member variables, the list of nodes and the number of nodes. Member functions include generating random Boolean networks, the simulation process, and an input/output user selection.

The Boolean node structure: The Boolean node class consists of the list of parents, truth tables, and a saved document of the values of the Boolean network as its member variables. The specific function is to evaluate the value of a node give the parent values.

The Boolean network format: The file format created for individual Boolean Networks is an input/output text file containing visual depiction of the truth tables, which can be easily translated into the network trajectories using the simulation algorithm.

Lastly, the specifications of the user have been determined to be the number of nodes the user wishes and the maximum amount of parents per node, determined by kmax.

Expected results: Ultimately, the final program will have user-friendly functions for easy usage. Generating random Boolean networks containing perturbations will be incorporated into the program. More importantly however, it is possible for a gene therapist to apply the data he or she has of a patient and use the program to find the most efficient result of which path to take. Also, a fourth portion of the program will be created. This section will contain a graphical procedure. In obtaining the trajectories from two genes, the growth and death of cells, the process will create a Flash generated visualization of the cell population over time.

Team Members: Jerry Yeh, Chris Smith, Ahra Cho
School Sponsor: Greg Marez