Global Snapshot of Protein Interaction Network -- A Percolation Based Approach

Abstract

In this paper, we study the large-scale protein interaction network of yeast uti lizing a stochastic method based upon percolation of random graphs. In order to find the global features of connectivities in the network, we introduce numeric al measures that quantify (1) how strongly a protein ties with the other parts o f the network and (2) how significantly an interaction contributes to the integr ity of the network. Our study shows that the distribution of essential proteins is distinct from the background in terms of global connectivities. This observ ation highlights a fundamental difference between the essential and the non-esse ntial proteins in the network. Furthermore, we find that the interaction data o btained from different experimental methods such as immunoprecipitation and two- hybrid techniques possess different characteristics. We discuss the biological implications of these observations.