 |
| |
 |
| |
 |
| |
 |
| |
 |
| |
 |
| |
 |
| Web site and design © 2008-2010 by G6G Consulting Group. All Rights Reserved. Most product content has been taken directly from manufacturer's web sites; other product content is assembled by G6G Consulting Group. G6G welcomes any corrections and/or comments. |
Category Cross-Omics>Agent-Based Modeling/Simulation/Tools Abstract ALC is a software tool for rule-based modeling that performs automated 'layer-based modeling' by following a model definition with a simple but advanced syntax. Conventional ‘mechanistic models’ can also be easily built. However, its focus is on simplifying and automating the construction of models that are according to the layer-based formalism. ALC supports the concepts of modules, rules and macrostates, which simplify the modeling procedure dramatically. The finished model is given as ready-to-run simulation files in the C MEX, MATLAB, Mathematica and Systems Biology Markup Language (SBML) formats. Output variables can be freely defined; their visualization is already included in the C MEX, MATLAB and Mathematica model files. A thorough consistency check is performed on the model definition. In the case of errors in the model definition detailed warnings and error messages are given that indicate the nature and location of the errors. ALC can be used offline or via a form on the ALC website. For offline use the freely available programming language Perl is required. ALC: Functionality and performance -- Simplification of the model generation - ALC converts model definitions given in a simple but advanced rule- based syntax to computational models in different formats, as well as documentation files. The main benefit of ALC is that it dramatically simplifies 'layer-based modeling' and reduces the risk of creating erroneous model equations. The assignment of the correction terms ci to dephosphorylation rates is one of the more difficult and error-prone steps in manual layer-based modeling. ALC performs these assignments automatically, such that errors in this step are avoided. This strengthens the analogy of layer-based modeling to conventional modeling and rule-based modeling as the reaction network within layers can now be defined using rules without considering correction terms. ALC also supports the usage of macrostates. This highly simplifies the definition of the signals between the layers (xi and xib) as well as the definition of conservation relations and the use of enzyme kinetics for rules and reactions. Portability of models - In many cases, it is comfortable to have the model in different formats. As an example, SBML is becoming the de facto standard for model representation in systems biology. Though by far Not all modeling and simulation projects use SBML, it is often desired to provide SBML models for publications. In many cases, the model equations in plain text are part of the manuscript or are provided as supplementary material. ALC automatically exports the model to several formats including SBML and provides the model equations in LATEX format as well as in plain text format. The manual format conversion of models is probably the major reason for errors in published model descriptions. Automation of this step, as provided by ALC, is Not only convenient but also lowers the probability of errors in the provided models. Application range of ALC - ALC is optimized for building layer-based models. It supports features of layer-based modeling that are Not present in conventional or rule- based modeling, and that are Not supported by other tools. ALC is also well suited for conventional modeling of ‘reaction networks’. Although the functionality of ALC is related to other tools that also allow for rule-based modeling, the application range is different. Note that rule-based conventional models can be built using ALC. However, due to limitations in the description of complexes, ALC is Not as well suited for this task as e.g. BioNetGen (see G6G Product Number 20466). ALC shows its optimal performance when building reduced modular models according to the layer-based approach. Computational aspects - The output files for small layer-based models are generated in less than one second on a desktop PC. The generation of larger layer-based models that can correspond to extremely large conventional models is also very rapid. As an example, the output files for the layer-based model of insulin signaling (214 ODEs) that replaces a conventional model with 1.5•108 ODEs are generated in about one second on a desktop PC. A few thousand Ordinary Differential Equations (ODEs) are usually generated within seconds. The model size in the offline version of ALC is only restricted by CPU time, disk space and memory capacity. The online version of ALC which is accessible via a form on the ALC website provides full functionality, but is restricted to model definitions that define No more than 500 species. This is done to keep the traffic and the processor load on the server at a reasonable size. The choice of the modeling method - An important question one has to answer at the beginning of the modeling process is that of which modeling method is the most suited for the considered system. A principle decision is necessary between stochastic and deterministic simulation. If stochastic simulation is chosen, it depends on the model size, as well as on the branching of the network and on the number of simulation runs as to whether on-the-fly or generate-first modeling is more suited. Note that layer-based modeling is also suited for stochastic simulation. In the following paragraph, the manufacturer gives a rule of thumb for choosing the appropriate deterministic modeling technique. The exact techniques should be used if it is possible and applicable. Whether one should use conventional modeling, rule-based modeling or exact model reduction simply depends on the size of the resulting model and on the number of necessary simulations. Additional assumptions (e.g. assuming a temporal order of processes) that are only made to keep the numbers of species and reactions low but have No physiological justification should be avoided. Such assumptions may lead to a nearly unpredictable approximation error.
modeling) is too large for the efficient simulation or parameter estimation, and if the exact reduced model is also too large or is difficult to generate, then layer-based modeling is a powerful possibility. In addition, the layer-based formalism is especially suited for comparing many ‘model variants’, as in many cases only a few equations in a single layer have to be changed. Note that rule-based modeling is also possible for layer-based models when using ALC. If the layer-based model is still too large for efficient simulation or parameter estimation, the layer-based model should be subjected to exact model reduction. In this case, the exact model reduction is performed on each layer separately. This combination of layer-based modeling and exact model reduction, results in the fewest ODEs, without the introduction of an additional approximation error, as introduced by layer-based modeling. System Requirements Web-based and contact manufacturer Manufacturer
Sandtorstr. 1 39106 Magdeburg Germany Manufacturer's Web Site http://layer.mpi-magdeburg.mpg.de/ Price Contact manufacturer G6G Product Number 20465 G6G Manufacturer Number 101743 |
| The G6G Directory of Omics and Intelligent Software |
 |
Search www.G6G-SoftwareDirectory.com
