Boyer, Frédéric (1,2);Morgat, Anne (1);Trilling, Laurent (2);Viari, Alain (1) - "ab initio'' metabolic pathway reconstruction

ECCB 2002 Poster sorted by: Author | Number

Next | Previous poster (in order of the view you have selected)

Title: "ab initio'' metabolic pathway reconstruction	P15
Boyer, Frédéric (1,2); Morgat, Anne (1); Trilling, Laurent (2); Viari, Alain (1) frederic.boyer@imag.fr, viari.alain@inrialpes.fr (1) INRIA Rhône-Alpes, (2) LSR-IMAG - Grenoble - France

Reconstructing metabolic pathways of fully sequenced organisms is a task of major importance. In order to help biologists in identifying the metabolic pathways of an organism, various tools have already been developed. A first approach is to rely on a database of already characterized metabolic pathways. Then for each pathway (or part of pathway) of the database, one try to find if it "occurs'' in the organism under study (to occur means, for instance, that the catalyzers are present and/or that the reactions are feasible)[1, 2, 3, 5]. By construction, this approach is not able to predict unknown or alternative metabolic pathways. A second more exploratory approach is the "ab initio metabolic pathway reconstruction'' that is the problem of predicting metabolic pathways without any other knowledge than a set of reactions (and the compounds they involve). By contrast, this approach allows to predict new (possibly unrealistic) pathways that should be further asserted.

We can informally state the problem of "ab initio metabolic pathway reconstruction'' as the following : Given a set of reactions and two compounds S and P, what are the possible combinations of successive reactions which lead to the syntesis of P from S. This is a computationaly difficult problem. Different algorithms have been proposed depending upon the constraints on the result (e.g no loss or gain of external compounds) or the nature of heuristics used [4, 6].

As an attempt to make the problem easier to handle, we reformulate the previous problem in a slightly different way : Given a set of reactions and two compounds S and P, what are the possible successive reactions which maximize the transfer of atoms from S to P.

To state this new problem, we have to consider reactions as rules describing the transfer of atoms between compounds. As reactions generally involve more than one substrate or product, we define a "matching'' as a relation which describe the transfer of atoms from one compound to another (see figure 1).

We then state the new problem as the one of finding all compositions of "matchings'' (see figure 2) which maximize the transfer of atoms from a compound S to a compound P.

On the poster we shall present the formal statement of the problem, the algorithms we have designed in order to resolve it and some preliminary results concerning the reconstruction of pathways.

Figure 1 : Extraction of a matching between Serine and Tryptophan from the reaction of EC number 4.2.1.20

Figure 2 : Composition of two matchings

(For the figures see the web representation of the poster abstract)

[1] Arvind K. Bansal. A framework of automated reconstruction of microbial metabolic pathways. In Proceedings of the IEEE International Symposium on Bio-informatics and Biomedical Engineering, 2000.
[2] Terry Gaasterland and Selkov Evgeni. Reconstruction of metabolic networks using incomplete information. In Proceedings of the International Conference on Intelligent Systems for Molecular Biology, 1995.
[3] Peter D. Karp, Markus Krummenacker, Suzanne Paley, and Jonathan Wagg. Integrated pathway-genome databases and their role in drug discovery. Trends in Biotechnology, 17(7), 1999.
[4] Robert Küffner, Ralf Zimmer, and Thomas Lengauer. Pathway analysis in metabolic databases via differential metabolic display (dmd). Bioinformatics, 16(9), 2000.
[5] Ross Overbeek, Niels Larsen, Natalia Maltsev, Gordon D. Pusch, and Evgeni Selkov. Bioinformatics, database and systems, chapter 3 WIT/WIT2 : Metabolic reconstruction system. Kluwer academic Publishers, 1999.
[6] Stefan Schuster, Fell David A., and Thomas Dandekar. A general definition of metabolic pathways useful for systematic organization and analysis of complex metabolic networks. Nature Biotechnology, 18, 2000.