GenYsisP toolbox

Category Cross-Omics>Pathway Analysis/Gene Regulatory Networks/Tools

Abstract The GenYsisP toolbox provides implicit methods for Stochastic Analysis of Gene Regulatory Networks.

The algorithms use Reduced Ordered Binary Decision Diagrams (ROBDDs) for implicit representation and traversal of Probabilistic Boolean Networks (PBNs) analysis of Gene Regulatory Networks (GRNs). Probabilistic Boolean Networks (PBNs) --

In a PBN, behavior of a gene can be described with multiple Boolean functions. Each function has a probability associated with it and there is at least ‘one function’ corresponding to each gene that can predict its expression as a function of the expressions of the input genes.

If all the genes have only one function then a PBN is similar to a Boolean Network (BN). Alternatively, a PBN can be seen as a set of BNs.

In that case each BN has a probability equal to the product of the probabilities associated with the Boolean functions of which it is composed.

Although most analyses on PBNs are done by looking at the latter description of a PBN, the manufacturer looks at the equivalent former description and provides a more suitable ‘mathematical model’ for the efficient analyses of ‘probabilistic gene regulatory networks’.

The manufacturers provide algorithms for the efficient ‘implicit representation’ of PBNs which enable analyses of large GRNs that were Not feasible earlier due to the corresponding computational complexity.

The new algorithms are currently available in the GenYsis-P toolbox (an extended version of the manufacturers GenYsis toolbox).

GenYsisP toolbox's Biological Motivation --

Dynamic analysis of gene regulatory networks (GRN) can be an advanced tool in understanding a cell differentiation process or the progression of a disease. One of the central features in such analyses is the identification of steady states (or the ‘attractors’) of the GRN.

If a given GRN represents the interactions between the genes/proteins participating in a cell differentiation process, then steady states may correspond to the ‘cell states’. Each cell state has a gene expression profile (i.e. a set of uniquely activated genes) that distinguishes it from other ‘cell states’.

Cell states may correspond to different functions of the cell such as proliferation, metabolism, apoptosis, etc. If a given GRN represents the pathways responsible for some diseases like cancer, then the cell states may even correspond to tumors.

The manufacturer envisions using PBNs for more advanced applications like experimental data analysis and on-demand drug therapies.

For example, given a PBN that is known to represent a ‘biological phenomenon’ and an ‘experimental dataset’ emerging from a new experiment on the same ‘biological system’, it would be interesting to associate a confidence measure with the data.

This would be particularly interesting if there is time series data, as the ‘dynamics of PBN’ can then be matched with the dynamics of the genes in the dataset.

Alternatively, given a PBN with unknown probabilities’on the Boolean functions, one could learn the probabilities from the experimental data.

This could be useful in highlighting the gene functions that are active in a given cell. Such analysis could be helpful for on-demand ‘drug therapies’ where treatment can be personalized to the patient under study.

Central to all these analyses is: computation of steady states, computation of the probability of a path from ‘one state’ of the network to another state and the identification of ‘key genes’ that should be perturbed to have a desired impact on the system.

All these computational tasks have been addressed in the literature.

However, current tools for PBNs use an ‘explicit representation’ and computation of the networks, restricting their application to networks having less than ten (10) genes. Even for small networks, these explicit techniques can Not detect the presence of ‘multiple attractors’.

GenYsisP toolbox's Commands and Functionalities --

1) Command name: help - Prints the list of available commands in GenYsisP on the terminal (computer screen).

2) Command name: randnet - Constructs a ‘synthetic gene regulatory network’.

3) Command name: readnet - Reads the gene regulatory network from the specified input file.

4) Command name: infonet - Prints information about the gene regulatory network in the current workspace.

5) Command name: printnet - Prints the gene regulatory network present in the current workspace. If an optional file name is given then the network is stored in the specified file else the network is printed on the terminal.

6) Command name: erasenet - Removes the network from the current workspace. Only one network can be present in the workspace at a time.

7) Command name: compinf - Computes the ‘influence matrix’ of the network.

8) Command name: printinf - Prints the ‘influence matrix’ of the network in a given file.

9) Command name: compsteady - Computes the ‘steady states’ (or ‘attractors’) of the network in the current workspace.

10) Command name: infosteady - Prints the information about the steady states (or ‘attractors’) on the terminal.

11) Command name: printstates - Prints the states in an ‘attractor’ to an output file.

12) Command name: probsteady - Computes steady state probability distribution of the ‘states in an attractor’. Prompts the user to skip ‘steady states’ whose size is very large and which may require unreasonable run time to construct the ‘state transition matrix’.

13) Command name: printsteady - Prints the ‘steady state’ probability distribution to an output file.

14) Command name: probpath - Computes the probability of going from the ‘source state’ to the ‘destination state’ along the ‘shortest paths’ in one or more steps.

15) Command name: run - Executes the commands in a given script file.

System Requirements

GenYsisP toolbox binary is currently available only for Linux.

GenYsisP has been compiled using gcc version 4.0, and requires ‘’ and ‘’ libraries. Also, contact the manufacturer for additional system requirements.


Manufacturer Web Site GenYsisP toolbox

Price Contact manufacturer.

G6G Abstract Number 20582

G6G Manufacturer Number 104185