| ECCB 2002 Poster sorted by: Author | Number Next | Previous poster (in order of the view you have selected) |
Title: Quantitative analysis of the large-scale organization of the protein-protein interaction network in yeast | P177 |
Wilhelm, Thomas; Nasheuer, Heinz-Peter; Suehnel, Juergen wilhelm@imb-jena.de, nasheuer@imb-jena.de, jsuehnel@imb-jena.de Institute of Molecular Biotechnology, Jena Centre for Bioinformatics, Beutenbergstr. 11, D-07745 Jena / Germany |
| Protein-protein interaction and metabolic networks have a large-scale organization that can neither be described by random characteristics nor by a regular lattice. Rather, they constitute small-world networks that are characterized by a preservation of local neighbourhood and by an logarithmic increase of the so-called network diameter with the number of vertices. In addition, they are usually assumed to be scale-free which means that the connectivity distribution follows a power law [1,2]. Such networks are known to be rather tolerant against a random disposal of vertices. On the other hand, highly linked protein nodes are essential for survival [3]. In a recent protein complex identification study it was stated that protein interaction networks are quite structured into subnets representing different cellular functions, such as cell cycle, protein synthesis, metabolism, etc. [4]. Finally, a comparative assessment of different methods for identifying protein-protein interactions has shown that different methods yield complementary results [5]. We have analyzed the connectivity distribution and the clustering characteristics of the yeast protein network by combining results from two different studies [4,6] thereby creating the most comprehensive data set currently available. The results are compared to the two-hybrid data obtained earlier [1,2]. As expected the network is not homogeneous but structured (small-world characteristics). However, contrary to the previous analyses we find with the larger data set that the connectivity distribution does not follow a power-law but is exponential. Quantitative information on protein connectivity is essential for an understanding of protein evolution and of information transfer pathways in cells. |
| [1] P. Uetz et al., Nature 2000, 403, 623-627. [2] I. Xenarios et al., Nucleic Acids Res. 2000, 28, 289-291. [3] H. Jeong et al., Nature 2001, 411, 41-42. [4] A.-C. Gavin et al., Nature 2002, 415, 141-147. [5] C. von Mering et al., Nature 2002, 417, 399-403. [6] Y. Ho et al., Nature 2002, 415, 180-183. |