Qualitatively Modelling Genetic Networks:
Petri Net Techniques and Tools
Principle Investigator: Richard Banks, PhD Student, School
of Computing Science, Newcastle University, UK (expected March
2009)
Project Supervisor: Dr Jason Steggles, School
of Computing Science, Newcastle University, UK
Key Contributors:
We would especially like to thank EPSRC for funding this
research, as well support from CISBAN and the Newcastle University
Systems Biology Resource Center.
Publications on this work are available.
Overview
This work aims to investigate, develop and evaluate the
application of Petri nets for qualitatively modelling and
analysing logical models of genetic regulatory networks. Petri nets are
well-established in the field of computing science, and
enjoy a number of attractive benefits, such as a wide range
of techniques and tools, which make them an invaluable resource for the biologist. However, there appears to be limited tool support
for systematically applying these Petri net techniques to
biological models.
A key result from this work is a suite of prototype tools
for aiding the biologist in model construction and analysis
with Petri nets. In particular, since most practical models
are too large to construct and interpret by hand, the
requirement for tool support is paramount.
Tools
A number of prototype tools to support our research have been developed. All tools are
written in Java, and have a straightforward GUI enabling the
user to start using them immediately. It
should be noted that these tools are still heavily in
development, and work is now ongoing to integrate these into
a single qualitative workbench for comprehensively studying genetic
regulatory networks.
GNaPN
Our Petri net construction tool which completely automates the construction of Petri net models from a logical network (e.g. Boolean or multi-valued network specified as state transition tables). GNaPN supports both the synchronous and asynchronous semantics of logical networks, systematically applies logic minimisation techniques to obtain a compact Petri net, can cope with non-determinism in situations when a model is partially understood, and can rapidly construct mutant models. Note GNaPN requires MVSIS
which has only been compiled and tested for Windows and Linux platforms. [Download] |
 |
ModelCreator
A simple utility for automatically creating the state transition
tables required by GNaPN. ModelCreator provides a
straightforward GUI for specifying the entities,
neighbourhoods and interactions present in a logical
regulatory network, and exports a series of state transition
tables describing this logic in a format understood by
GNaPN. [Download] |
 |
STGRefiner
A tool for dynamically and interactively constructing
realistic asynchronous models. STGRefiner takes as input the
signal transition graphs produced by GNaPN and automatically
identifies all missing reaction rate information. Given such
information, the user is then able to observe the precise
points in the model where more information is required, and
can automatically resolve this by inputting relative
reaction rates. Backtracking is also
possible, making the rapid exploration of different
experimental hypotheses straightforward. [Download] |
 |
Models
A selection of model files (specified as state transition tables) are given below. These are the files that can be produced by ModelCreator and used as input to GNaPN:
Links
Below are some links to existing tools which can be used to analyse our Petri net models:
- PEP - integrated Petri net environment.
- Punf - Parallel Petri net unfolder for efficient model checking.
- Clp - a linear programming model checker which works either a separate utility or as part of PEP.
- Petrify - a tool for the synthesis of Petri nets from asynchronous specifications. In particular, Petrify can be used to analyse and refine the signal transition graphs produced by GNaPN.
- Visual STG Lab - an improved visual front end to Petrify.
Note as GNaPN can export Petri nets to PNML, this also allows them to be analysed using a wide range of other tools (for example, see here).
© 2009 Richard Banks, School of Computing Science, Newcastle University, UK |
