Home
Welcome to the kboolnet wiki! You'll find documentation for the various scripts and functions of the Kufareva Lab's rxncon pipeline here.
kboolnet is an R package and set of associated R and Python scripts created for the automated verification, validation, and visualization (VVV) of rxncon models.
The rxncon formalism, developed by Marcus Krantz and Edda Klipp at Humboldt University, represents a signaling network as a bipartite directed graph of nodes of two types, states and reactions. State nodes represent the state of proteins at the domain/residue level. These states are connected to each other by reaction nodes, which represent uni- or bi-molecular transformations which consume a state(s) and produce another state(s). State nodes are also connected to reaction nodes by contingency edges, which indicate how a particular state regulates a reaction.
A rxncon model's specification consists of a list of reactions, a list of contingencies upon these reactions, a list of reaction types, and a list of modification types. These are typically encoded as spreadsheets in the Microsoft XLS/XLSX format; a simplified example of such a spreadsheet may be downloaded here.
Once written, a rxncon model can be converted into an executable Boolean network. The kboolnet toolkit can be used to verify, validate, and visualize the behavior of these Boolean networks and the associated rxncon model. The toolkit also provides an interface with Google Drive, allowing for the cloud storage and collaborative editing of rxncon models.
Verification is the process of ensuring a model is internally consistent and is carried out by the VerifyModel.R script. VerifyModel.R formally verifies a model by measuring both its ability to respond to repeated rounds of ligand treatment and the consistency between Boolean steady states representing the same biological state.
Validation is the process of assessing a model's agreement with experimental data and expectations for its behavior. Three scripts are available for model validation:
- TruthTable.R - Simulate a model under all possible combinations of given activators/inhibitors and plot their effect on selected output nodes.
- SensitivityAnalysis.R - Measure effect of individually inhibiting each node in the model on the effect a ligand has on a set of output nodes
- ScoreNet.R - Simulate a model under perturbations encoded in a database of experimental data and compare simulation results to real-world observations
- PlotPath.R - Plot the trajectory of a Boolean network
- PlotPathComparison.R - Overlay two Boolean network trajectories, showing their similarities and differences
- AnimatePath.R - Animate the regulatory graph of a rxncon model based on a Boolean network trajectory
- PlotModules.R - Make regulatory graphs for each module in a rxncon model
Refer to Installation for installation instructions, Running scripts for a guide on how to use the kboolnet scripts, and FAQ and possible errors for solutions to problems frequently encountered during installation and use.
The kboolnet package and associated scripts were developed by Willow Carretero Chavez, Marcus Krantz, Edda Klipp, and Irina Kufareva. Many thanks are given to Alexis Lona and other members of the Kufareva Lab for assistance in testing kboolnet.