The amount of fast growing available protein-protein interactions has provided unprecedented opportunities for understanding the mystery of biology. The Yeast as one of model organisms has many different protein-protein interaction (PPI) datasets, but there is no common approach for integration these datasets to analyses the topological structure of PPI network. In this study, we reconstructed a new weighted PPI network by integrating high-confidence PPI datasets in a way that the debate of the datasets has great influence to results.

Analysis the Topological Structure of Reconstructed High-confidence Yeast PPI Network

The centrality-lethality rule reveals that topological prominence of a protein indicates its functional importance. But are all hub proteins in PPI network essential? Using P-value to weight PPI networks, we classified hub proteins into two types: essential hub and nonessential hub. The weighted centrality measure of weighted PPI networks showed that the weighted method can give greater support for functional characterization of hub proteins. We found that essential hubs are vital and they tend to be intra-modular hubs, while nonessential hubs are useful in maintaining the overall network connectivity and tend to be inter-modular hubs.

Functional characterization of hub proteins in weighted yeast protein-protein interaction networks

Centrality-lethality rule demonstrates that proteins with high degree in the protein-protein interaction (PPI) network are more likely to be essential than those selected by chance. Nevertheless, in PPI network, there are still many proteins with high degree are not essential (false hubs). Comparing the difference between false hubs and essential hubs from topological perspective, such as centrality measures, damage, clustering coefficient and interactions between neighbors, there are distinguishable significant properties of false hubs. False hubs have lower sub graph centrality and less interactions between neighbors, which could explain why they are not essential in some extent.

http://ieeexplore.ieee.org/iel5/6401247/6406849/06406880.pdf

From the Topological View to Explain Why False Hub is Not Essential

The reason for centrality-lethality is still an intriguing question. The idea of party and date hub proteins has been influential in the study of protein-protein interaction networks. But the validity of the party/date hub distinction has been debated. Using average Pearson correlation coefficient in cell cycle, stress response and DNA damage, we confirmed the distinction of E-party hub and E-date hub proteins. The results also support the classification of hub proteins into party hub and date hub proteins.

Understanding centrality-lethality rule from the distinction of essential party/date hub proteins in yeast protein-protein interaction networks

Detecting protein complexes is an important way to discover the relationship between network topological structure and its functional features in protein-protein interaction (PPI) network. The spectral clustering method is a popular approach. However, how to select its optimal Laplacian matrix is still an open problem. Here, we analyzed the performances of three graph Laplacian matrices (unnormalized symmetric graph Laplacians,, normalized symmetric graph Laplacians and normalized random walk graph Laplacians, respectively) in yeast PPI network. The comparison shows that the performances of unnormalized and normalized symmetric graph Laplacian matrices are similar, and they are better than that of normalized random walk graph Laplacian matrix. It is helpful to choose proper graph Laplacian matrix for PPI networks’ analysis.

Performance Analysis of Graph Laplacian Matrices in Detecting Protein Complexes

One of the important problems in system biology is to discover the relationship between topological properties and functional features of proteins in protein-protein interaction (PPI) networks. There are many essential-nonhub proteins and lots of nonessential-nonhub proteins in the PPI network. Both essential-nonhub proteins and nonessential-nonhub proteins are low-connectivity proteins, but they have different lethality. In order to explain why nonessential-nonhub proteins are not essential, we compare them with essential-nonhub proteins from topological view. The comparison results show that there are statistical differences between nonessential-nonhub proteins and essential-nonhub proteins in centrality measures and clustering coefficient.

http://ieeexplore.ieee.org/iel5/6331799/6340697/06340764.pdf

Dong Yun-yuan

Papers

Functional characterization of hub proteins in weighted yeast protein-protein interaction networks

From the Topological View to Explain Why False Hub is Not Essential

Understanding centrality-lethality rule from the distinction of essential party/date hub proteins in yeast protein-protein interaction networks

Performance Analysis of Graph Laplacian Matrices in Detecting Protein Complexes

The topological features of nonessential-nonhub proteins in the protein-protein interaction network