Limits...
Protein interaction networks--more than mere modules.

Pinkert S, Schultz J, Reichardt J - PLoS Comput. Biol. (2010)

Bottom Line: Such grouping may well result in cohesive modules again, but only if the network structure actually supports this.When mapping experimental methods onto the groups, we identified profound differences in the coverage suggesting that our method is able to capture experimental bias in the data, too.For example yeast-two-hybrid data were highly overrepresented in one particular group.

View Article: PubMed Central - PubMed

Affiliation: Department of Bioinformatics, Biocenter, University of Würzburg, Würzburg, Germany.

ABSTRACT
It is widely believed that the modular organization of cellular function is reflected in a modular structure of molecular networks. A common view is that a "module" in a network is a cohesively linked group of nodes, densely connected internally and sparsely interacting with the rest of the network. Many algorithms try to identify functional modules in protein-interaction networks (PIN) by searching for such cohesive groups of proteins. Here, we present an alternative approach independent of any prior definition of what actually constitutes a "module". In a self-consistent manner, proteins are grouped into "functional roles" if they interact in similar ways with other proteins according to their functional roles. Such grouping may well result in cohesive modules again, but only if the network structure actually supports this. We applied our method to the PIN from the Human Protein Reference Database (HPRD) and found that a representation of the network in terms of cohesive modules, at least on a global scale, does not optimally represent the network's structure because it focuses on finding independent groups of proteins. In contrast, a decomposition into functional roles is able to depict the structure much better as it also takes into account the interdependencies between roles and even allows groupings based on the absence of interactions between proteins in the same functional role. This, for example, is the case for transmembrane proteins, which could never be recognized as a cohesive group of nodes in a PIN. When mapping experimental methods onto the groups, we identified profound differences in the coverage suggesting that our method is able to capture experimental bias in the data, too. For example yeast-two-hybrid data were highly overrepresented in one particular group. Thus, there is more structure in protein-interaction networks than cohesive modules alone and we believe this finding can significantly improve automated function prediction algorithms.

Show MeSH

Related in: MedlinePlus

Block assignment in a functional role model with 100 classes.Adjacency matrix with rows and columns ordered according to assignment of proteins into classes. Color code and size of dots representing matrix entries as in figure 5. Only classes containing more than 100 proteins are labeled for better readability. Two details from the corresponding image graph exemplifying the kinds of structures found by the algorithm.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2813263&req=5

pcbi-1000659-g006: Block assignment in a functional role model with 100 classes.Adjacency matrix with rows and columns ordered according to assignment of proteins into classes. Color code and size of dots representing matrix entries as in figure 5. Only classes containing more than 100 proteins are labeled for better readability. Two details from the corresponding image graph exemplifying the kinds of structures found by the algorithm.

Mentions: Next, we consider a non-diagonal model with classes. Figure 6 shows the adjacency matrix of the network with rows and columns ordered according to the assignment found by the algorithm. The entire 100 node image graph is connected but too complex to be discussed within the scope of this paper. Instead, we focus on two small subgraphs, as shown in figure 6, which exemplify two typical substructures in the network that could not be discerned by methods focusing on cohesive modules alone, and discuss their biological interpretation in greater detail. Note that in figure 6 we only show the classes and links between classes discussed below and that the majority of classes shown have additional connections with the rest of the image graph.


Protein interaction networks--more than mere modules.

Pinkert S, Schultz J, Reichardt J - PLoS Comput. Biol. (2010)

Block assignment in a functional role model with 100 classes.Adjacency matrix with rows and columns ordered according to assignment of proteins into classes. Color code and size of dots representing matrix entries as in figure 5. Only classes containing more than 100 proteins are labeled for better readability. Two details from the corresponding image graph exemplifying the kinds of structures found by the algorithm.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2813263&req=5

pcbi-1000659-g006: Block assignment in a functional role model with 100 classes.Adjacency matrix with rows and columns ordered according to assignment of proteins into classes. Color code and size of dots representing matrix entries as in figure 5. Only classes containing more than 100 proteins are labeled for better readability. Two details from the corresponding image graph exemplifying the kinds of structures found by the algorithm.
Mentions: Next, we consider a non-diagonal model with classes. Figure 6 shows the adjacency matrix of the network with rows and columns ordered according to the assignment found by the algorithm. The entire 100 node image graph is connected but too complex to be discussed within the scope of this paper. Instead, we focus on two small subgraphs, as shown in figure 6, which exemplify two typical substructures in the network that could not be discerned by methods focusing on cohesive modules alone, and discuss their biological interpretation in greater detail. Note that in figure 6 we only show the classes and links between classes discussed below and that the majority of classes shown have additional connections with the rest of the image graph.

Bottom Line: Such grouping may well result in cohesive modules again, but only if the network structure actually supports this.When mapping experimental methods onto the groups, we identified profound differences in the coverage suggesting that our method is able to capture experimental bias in the data, too.For example yeast-two-hybrid data were highly overrepresented in one particular group.

View Article: PubMed Central - PubMed

Affiliation: Department of Bioinformatics, Biocenter, University of Würzburg, Würzburg, Germany.

ABSTRACT
It is widely believed that the modular organization of cellular function is reflected in a modular structure of molecular networks. A common view is that a "module" in a network is a cohesively linked group of nodes, densely connected internally and sparsely interacting with the rest of the network. Many algorithms try to identify functional modules in protein-interaction networks (PIN) by searching for such cohesive groups of proteins. Here, we present an alternative approach independent of any prior definition of what actually constitutes a "module". In a self-consistent manner, proteins are grouped into "functional roles" if they interact in similar ways with other proteins according to their functional roles. Such grouping may well result in cohesive modules again, but only if the network structure actually supports this. We applied our method to the PIN from the Human Protein Reference Database (HPRD) and found that a representation of the network in terms of cohesive modules, at least on a global scale, does not optimally represent the network's structure because it focuses on finding independent groups of proteins. In contrast, a decomposition into functional roles is able to depict the structure much better as it also takes into account the interdependencies between roles and even allows groupings based on the absence of interactions between proteins in the same functional role. This, for example, is the case for transmembrane proteins, which could never be recognized as a cohesive group of nodes in a PIN. When mapping experimental methods onto the groups, we identified profound differences in the coverage suggesting that our method is able to capture experimental bias in the data, too. For example yeast-two-hybrid data were highly overrepresented in one particular group. Thus, there is more structure in protein-interaction networks than cohesive modules alone and we believe this finding can significantly improve automated function prediction algorithms.

Show MeSH
Related in: MedlinePlus