| Home |
| About |
| People |
| Links |
| SysBio? |
| Research |
| Projects |
| Pubs |
| Suppl. |
| Education |
| Courses |
| Theses |
| Info |
| Publicity |
| Talks |
| Jobs |
| Impressum |
|
|
A Glossary for Systems Biology
Contents
Index - Glossary -
Glossary
A Closer
Model
Submit - Comment
Modularity
Modularity is a design concept in engineering, enabling engineers to build complex systems out of simpler modules, which can be tested independently before being integrated. As an additional benefit, such modules can be used in more than one system, saving development time and increasing quality. Modules also make maintenance of such systems much easier. These two aspects increase security with such complex systems, through ease of testing during design and by facilitating failure detection and correction during operation.
Modularity seems to have applications in biology as well. There are metabolic units which can be found in different organisms [39] and homologies are used to compare the genomes of different organisms in order to gain insights into development, physiology and evolution [24]. Some researchers hope that such comparative research could reveal a ''modular framework'' for biology in which subsystems of complex molecular networks can be treated as ``functional units that perform identifiable tasks perhaps even able to be characterized in familiar engineering terms'' [39].
It might become evident in the future that this is not the best representation of how biological systems are constructed. Nevertheless, such a modular concept has advantages; for example it would be a powerful aid for research, enabling researchers to use the full range of tools that have been developed for this approach in engineering. It could also make modeling and handling of models a lot more effective.
There is hope, though, that it will be more than that, because the reason for engineers to chose this concept is one that also applies in biology: efficiency. Even though a lot of biological systems are widely agreed to be far from optimal (i.e. most economical) in their construction or operation, this is usually due to evolutionary development or the need for robustness or flexibility, and can in itself be seen as a kind of optimization. Regardless of these considerations, it is also widely agreed that evolution, wherever possible, tries to change in the direction of more efficiency. So if modularity can be seen as a way to achieve that, there is reason to hope that the concept of modularity is a real factor in biological systems as well.
If this proves to be the case it might open up completely new ways of describing biological systems, their components and their evolutionary development [34].
Apart from the advantages mentioned before this would also coincide nicely with the concept of systems in systems theory (system, modularity), where scientists think in terms of classes of systems, defined by a certain set of common characteristics, which can be handled by a common set of methods.
Some good biological examples are given in [39]. These include the module for bacterial chemotaxis, which has been compared to mechanisms for phototaxis in other texts, according to [39]. A number of different functions have been attributed to another module, the MAP kinase cascade. Among those mentioned are ``amplifier for an upstream receptor/ligand binding event'' and ``switch providing an almost threshold-like input/output response'', as well as exact adaptation and feedback control [39]. In fact it might be argued that this module can fulfill all these functions depending on its interconnection with other modules, that is, that it can be tuned through a set of parameters.
It remains to be seen whether the concept of modularity really is a governing principle in the construction of biological systems and can be exploited as detailed above. In that case it would help to solve a number of problems that have hampered progress in interpreting experimental data. One of those is scale, or level of detail (also level of abstraction), to be used in a model for in-silico research.
One big problem of current models is that they already are much too cumbersome for manual analysis, cannot be simulated efficiently because of their size and complexity, and do not lend themselves to easy automatic segmentation (yet). Manual segmentation, on the other hand, is a highly complicated task that ties up a lot of human resources. Often the size of the model effectively prohibits manual segmentation completely.
These problems will only get worse when new, more detailed experimental data are incorporated into the models. Should there be a natural structure provided by functional modules, that might solve many of these problems. There would be smaller units for manual analysis; segmentation would not be necessary anymore (at least not on this scale); and the level of detail to be used in a simulation might easily suggest itself by evaluating the influences of lower-level modules on the modular level of interest and then neglecting - or at least simplifying - simulation of those modules which exert only weak influences (model).
Next: Glossary Up: A Closer Previous: Model Contents Index
Glossary Submit Comment
Universität Magdeburg
University of Liège