Chen, Ming - Metabolic Pathway Alignment and Alternative Pathways Identification

ECCB 2002 Poster sorted by: Author | Number

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

Title: Metabolic Pathway Alignment and Alternative Pathways Identification	P22
Chen, Ming mchen@techfak.uni-bielefeld.de Technische Fakultaet, Universitaet Bielefeld, Germany

With the technological breakthrought in molecular biology and rapid development of bioinformatics, more and more experimental data both on gene and cell levels are systematically collected and stored in specific databases which are also available to public via internet. In the so-called post-genomic era, researchers are willing to find ways to deal with all these exponentially increasing data to better understand the mystique of life. It is thus not surprising that the development of computational algorithms to predict metabolism function from gene, amino acid sequences and metabolic networks is now a core aim of bioinformatics, of which identification and analysis of metabolic pathways is a promising field as the completion of a long series of genomes and the accumulation knowledge of metabolism have made the comparison of complete metabolic pathways possible. Moreover, studying on metabolic pathways is essential to do research forward from genotype to phenotype and to work on metabolic engineering.
Metabolic pathway is defined in the literature in various ways with varying degrees of formality. In biological papers, pathways are often named after key molecules (e.g. "glycolysis", "pentose phosphate pathway", "urea cycle"). Typical metabolic pathways are given by the wall chart of Boehringer Mannheim and KEGG, which have been verified with a number of printed and on-line sources. In that context, pathway refers to a set of paths from substrate to product. However, the definition is quite ambiguous there are always interactions among pathways. Now, the question is how to define a boundary for a pathway under these circumstances that the biology processes are so interacted and actually there is no such clear boundary between two pathways; and how to classify those metabolites that are exclusive to the known pathways. We are motivated to give one more generalized definition of metabolic pathway.
One prevailing definition of a metabolic pathway is a directed reaction graph with substrates as vertices and arcs denoting enzymatic reactions [2]. A similar definition for a metabolic pathway is given as P=G(M,E,A), an unrepeatable, irreversible sequence of a series of vertices and arcs leading from a molecule vertex labeled as substrate via a molecule vertex labeled as enzyme to a molecule vertex labeled as product. A metabolic pathway consists of substrates and products (represented as vertices M); enzymes (represented as vertices E) and arcs from substrates to enzymes and from enzymes to products (represented as arcs A). Obviously, a metabolic pathway is a special part of complex network of reactants, products and enzymes with multiple interconnections representing reactions and regulation. A pathway does not operate in isolation from other pathways, and thus cannot be considered as a distinct physical entity.
Alignment as one of the most powerful methods to comparatively analysis the relationship between two sequences has been widely investigated in bioinformatics field. Researches on gene sequence alignment are so far intensively done and many applications and tools, such as BLAST and FASTA are developed to further understand the biological homology and estimate evolutionary distance. Recently the emphasis of research efforts begins to turn back from gene sequences to metabolic pathways. Metabolic pathway alignment is of importance to study biology evolution, pharmacological targets and other biotechnological applications [1] such as metabolic engineering, metabolism computation. Several approaches of metabolic pathway alignment are already made in the past years. Forst C.V. [2] [3] extended the DNA sequence alignment methods to define distances between metabolic pathways by combining sequence information of involved genes. Dandekar et al. [1] compared glycoslysis, Entner-Doudoroff pathway and pyruvate processing in 17 organisms based on the genomic and metabolic pathway data by aligning specific pathway related enzyme-encoding genes on the genomes. Tohsato Y. et al. [4][5] proposed a multiple (local) alignment algorithm utilizing information content that was extended to symbols having a hierarchical structure EC numbers. We proposed a pairwise pathway alignment strategy, based on substrate-enzyme-product pattern comparison, to analysis the similarities among metabolic pathways.
According to our definition of metabolic pathway, metabolic pathways are partition of metabolic networks which contain both metabolites (substrates & products) and enzymes, an enzyme makes its substrate(s) and product(s) as a reaction unit. A pathway begins with a substrate and ends up with a product. The alignment algorithm is based on likelihood calculations; it calculates the differences of metabolites from both ends of pathway as well as the enzymes shared with the same substrate-product pattern. Enzyme is expressed with EC number which shows 4 level hierarchy. It is compared by scoring the hierarchical similarity of their EC numbers. Alignment score indicates the degree of similarity of two metabolic pathways. Metabolic pathways are called alternative pathways when they share the same start substrate and end product. As a result, it is possible to pick up two arbitrary pathways and align them to identify whether they are alternative pathways or partially are. This method differs from classic alternative pathway finding based on Dijkstra's algorithm which is used by some well known metabolic pathway databases such as KEGG, aMAZE and WIT.
In addition, a web-based interface is being developed to implement the metabolic pathway alignment algorithm. The two-pathway alignment is updated and available for public at http://apogonidae.techfak.uni-bielefeld.de/mchen/PathAligner/. Aligning pathways through the database is still under-construction because of the development of metabolic pathway database.

[1] Dandekar T., Schuster S., Snel B. et al. (1999) Pathway alignment: application to the comparative analysis of glycolytic enzymes. Biochem J. 1: 115-24.
[2] Forst C.V. and Schulten, K. (1999) Evolution of metabolisms: a new method for the comparison of metabolic pathways using genomics information. J. Comput. Biol. 6: 343-360.
[3] Forst C.V. and Schulten K. (2001) Phylogenetic Analysis of Metabolic Pathways. J. Mol. Evol. 52: 471-489.
[4] Tohsato Y., Matsuda H. and Hashimot A. (2000) A Multiple Alignment Algorithm for Metabolic Pathway Analysis using Enzyme Hierarchy. Proc. 8th International Conference on Intelligent Systems for Molecular Biology (ISMB 2000) 376-383.
[5] Tohsato Y., Matsuda H. and Hashimot A. (2000) An Application of a Pathway Alignment Method to the Analysis of Amino Acid Biosynthesis. Genome Informatics. 11: 284-285.