Category Cross-Omics>Pathway Analysis/Tools

Abstract GINsim (Gene Interaction Network simulation) is a software tool for the modeling and simulation of genetic regulatory networks.

GINsim consists of a simulator of qualitative models of genetic regulatory networks based on a discrete, logical formalism.

GINsim allows the user to specify a model of a genetic regulatory network in terms of asynchronous, multi-valued logical functions, and to simulate and/or analyze its qualitative dynamical behavior.

GINsim is a tool supporting the definition, the simulation and the analysis of regulatory graphs, based on a 'multilevel logical formalism'.

This formalism leans on the definition of two types of graphs: 'regulatory graphs', which model regulatory networks, and 'state transition graphs', which represent their dynamical behavior, for given initial state(s) and under specific updating assumptions.

Defining a regulatory network model -- In logical formalism, a regulatory network is modeled as a 'regulatory graph'.

In this graph, nodes represent genes or regulatory products, and arcs represent interactions between genes. Modeling attributes determine the dynamic behavior of the network.

GINsim allows you to define interactively such models through a dedicated graphical interface. A graph editor allows the addition/removal of genes and interactions.

And, the property panel at the bottom of the main GINsim window enables the specification of the 'modeling attributes'.

Drawing the regulatory graph -- The first step in the definition of a 'regulatory model', is the definition of the graph per se.

Genes and interactions can be interactively added, selected, moved or deleted, depending on the current 'editing mode'.

Available editing modes include:

1) Default editing mode - allows you to select and move objects;

2) Gene insertion mode - when selected, clicking on the graph panel adds a new gene;

3) Interaction insertion mode - when selected, interactions are added by dragging from one gene to another. The interactions must be complemented by the definition of the 'logical parameters' for the target variable (see below).

These three buttons allow you to add different kinds of interactions: activation, inhibitions or undefined.

4) Deletion option (mode) - selected items are deleted; 5) Arc modification mode - when selected, intermediate points can be added or removed.

Entering parameters for genes and interactions -- Each gene has a maximal and a basal expression level. The 'basal level' corresponds to the situation where No incoming interaction is active.

GINsim assigns a default maximal level of 1 and a basal level of 0 to newly created genes.

When a gene is selected, the 'modeling attributes' feature allows you to edit properties of this gene. Name (optional), Id (unique identifier) and expression levels can be edited.

Interactions operate for a given range of activity levels based on their source, whenever necessary you can define these ranges for each interaction, [per default, the interval is set to (1-max), where "max" is the maximal expression level of the source].

For a selected interaction, these values can be specified through the modeling attributes area. Each interaction has a source and a target gene. An interaction is active for an encompassed range of activity levels based on its source gene.

Editing Logical Parameters -- Logical parameters are attached to each gene, to describe its dynamical behavior.

Each logical parameter is defined as a set of incoming interactions and a target value. It denotes that, during the simulation, when:

a) the specified incoming interactions are all active (i.e. the expression level of their source genes are in the related activity ranges);

b) and none of the other incoming interactions is active;

then, the ‘expression level’ of the considered gene tends toward the value attached to the logical parameter.

Per default, all parameters are considered to be zero. Consequently, only the non-zero parameters have to be explicitly defined.

Saving the graph -- GINsim allows you to save, open or export files. GINsim uses a dedicated XML format: GINML.

GINML extends GXL (Graph eXchange Language - GXL is an XML- based standard exchange format for sharing data between tools. Formally, GXL represents typed, attributed, directed, and ordered graphs which are extended to represent hyper-graphs and hierarchical graphs).

The GINML extension comprises new attributes and sub-elements for elements 'node and edge', and an additional element called 'parameter'.

Running a simulation -- Once a regulatory model has been defined, a simulation can be launched.

The simulation computes a 'state transition graph', representing the dynamical behavior of the model.

The nodes of this graph, also called states of the system, are vectors giving the ‘expression levels’ of all genes of the model.

At a given state, a specific logical parameter can be associated with each regulatory component.

If the value of this parameter is smaller or greater than that of the concentration/activity level of the corresponding component, this level will tend to decrease or increase, respectively.

Otherwise (when the parameter value and the corresponding component level are equal), the component will tend to keep its current value.

Building a ‘state transition graph’ involves the selection of a strategy for its construction, the specification of initial state(s) as well as optional size constraints.

Initial state(s) -- The simulation can consider all possible states as initial states (resulting in a full state transition graph), or use only a specific subset, defined in a table.

Each line of this table describes one or several initial states, by means of a set of allowed values for each gene of the system.

Construction strategy -- When several genes are called to update their expression levels at the same time, several strategies can be applied.

In GINsim, the following construction strategies are implemented:

1) Synchronous - all the updating calls are executed simultaneously. This strategy is generally considered unrealistic, but it is still useful in some cases.

2) Asynchronous - only one updating call is executed at each step, but all possible trajectories are generated in the resulting state transition graph. This is the default strategy.

3) Priority classes - this is a mixed mode where updating calls are grouped in priority classes.

Priority classes -- GINsim allows the user to group components into different classes, and to assign a priority level to each of these classes.

In case of concurrent transition calls, GINsim first updates the gene(s) belonging to the class with the highest ranking.

For each regulatory component class, the user can further specify the desired updating assumption, which then determines the treatment of concurrent transition calls inside that class.

When several classes have the same ranking, concurrent transitions are treated under an asynchronous assumption (No priority).

By default, all transitions belong to the same asynchronous class. Running the simulation without further configuration is thus the same as using the asynchronous assumption.

Transition blocking -- Blocking transition is putting some restrictions on the evolution of the expression level of genes.

The transition blocking feature allows you to specify minimal and maximal expression level for each gene.

Note: Transition blocking is especially useful to simulate simple or multiple mutations (knockouts or ectopic gene expression).

Analysis Tools --

Strongly Connected Component - GINsim enables the identification of the Strongly Connected Components graph (SCC graph) of an existing graph.

The SCC graph is an acyclic graph based on the original graph: each node of the SCC graph is a cycle or a set of intertwined cycles of the original graph.

Path Finder - It is often important to find a path between different states of the state transition graph. The "search path" action feature allows you to find a path between different states of the state transition graph.

Animation Tool - The animation tool allows you to select a path interactively. It highlights genes of the regulatory graph according to the selected state.

System Requirements

GINsim is a Java application and thus should run on any system offering a Java virtual machine (version 1.4 or later).

Other system requirements depend heavily on model size.

To work comfortably, a computer with at least 512 Mb of RAM is required. A screen resolution of at least 1024x768 is recommended.


Manufacturer Web Site GINsim

Price Freely available for academics, please contact us for other uses.

G6G Abstract Number 20333

G6G Manufacturer Number 102866