One of the main tools used in systems biology is genome-scale metabolic models. These models contain, in principle, all capabilities in the metabolism of an organism (all chemical reactions that can take place inside its cell or cells). These models are basically a list of all chemical reactions catalyzed by the enzymes found in an organism's genome.
There are different ways to study such models. The most traditional one is the flux balance analysis (FBA), defining an objective reaction an optimizing the flux in ever reaction of the organism in order to maximize the flux in this objective. This kind of analysis has many applications for studying conditions for optimal growth of an organism or its production capabilities for some metabolite of interest. Another possible way to study metabolic models is to do a theoretical study of its underlying networks. These models represent rich structures from the graph theoretical point of view and many tools from this theory can be applied here.
Although there are several different software available in the market for the aforementioned analyzes, the majority are desktop applications with limited resources, or proprietary software for which licenses are very expensive. PyNetMet, on the other hand is a set of fully programmable classes available for python for free.
PyNetMet has been develop inside the Intertech Group. The Intertech group is an interdisciplinary research group formed by mathematicians, physicists, engineers, biologists and psychologists.
Other researches that contributed in the development and testing of PyNetMet are A. Montagud, R. J. Infante, M. Siurana, D. Fuente, J. Triana, P. Fernandez de Cordoba and J. Urchueguia.
Details about the classes, its methods and attributes can be found in the PyNetMet manual which come in the linux distribution of the PyNetMet package or can be downloaded directly from here. The manual also has examples of uses for the different classes. How to create networks and metabolic models directly from the interpreter, how to parse and translate models from different file formats, etc.
Examples on how to use PyNetMet for different analyzes over a real metabolic model can be found here.
PyNetMet is in its first version. As with any other software, a few bugs have been already identified and will be fixed in the next version. The known bugs of the current version, and possible walkarounds to avoid them are in the table below.
| Class | Method | Bug | Possible Solution |
|---|---|---|---|
| Enzyme | __add__ | The method is not conserving issues from its parent reactions. | If important, check for issues before summing enzymes. |
| Metabolism | add_reacs | Is not recognizing transport reactions properly. It also fails to update other Metabolism attributes. | After using this method, run the calcs method in order to properly update all the Metabolism object. It takes longer, but works. |
| Metabolism | pop | Has same similar problems as add_reacs. | Same solution as for add_reacs. |
If you find any other bug, please report it to the author.
The next version of PyNetMet will have these bugs corrected. It will also contain new functions for generating different kinds of random networks, and the Metabolism and Network classes will be more efficient (faster) for some purposes.