Modeling interactome: scale-free or geometric?

Abstract

Supplementary information is available at http://www.cs.utoronto.ca/~juris/data/data/ppiGRG04/

Network structure and the biology of populations

“…Not the least of these is that many of the observed degree distributions for intracellular signalling, although roughly linear on a log-log plot (i.e. a power law) over much of their range, are in fact better fit with an exponential or other degree distribution [20,21]. Another problem, only recently pointed out [22], is that many of these networks are in fact samples (often rather small samples of 10% to 20%) of the full network in question.…”

Graphlet decomposition: framework, algorithms, and applications

“…Much of the work in this vein focused on analyzing triadic tendencies as important structural features of social networks (e.g., transitivity or triadic closure) as well as analyzing triadic configurations as the basis for various social network theories (e.g., social balance, strength of weak ties, stability of ties, or trust (Granovetter, 1983)). In biology (Pržulj et al, 2004;Milenkoviae and Pržulj, 2008), graphlets were widely used for protein function prediction (Shervashidze et al, 2009), network alignment (Milenković, Ng, Hayes and Pržulj, 2010), and phylogeny (Kuchaiev, Milenković, Memišević, Hayes and Pržulj, 2010) to name a few. More recently, there has been an increased interest in exploring the role of graphlet analysis in computer networking (Feldman and Shavitt, 2008;Hales and Arteconi, 2008;Becchetti, Boldi, Castillo and Gionis, 2008) (e.g., for web spam detection, analysis of peer-to-peer protocols and Internet AS graphs), chemoinformatics (Ralaivola, Swamidass, Saigo and Baldi, 2005;Kashima, Saigo, Hattori and Tsuda, 2010), image segmentation (Zhang, Song, Liu, Liu, Bu and Chen, 2013), among others (Zhang, Han, Yang, Song, Yan and Tian, 2013).…”

“…These patterns are called graphlets (Pržulj, Corneil and Jurisica, 2004). Graphlets (also known as motifs (Milo, Shen-Orr, Itzkovitz, Kashtan, Chklovskii and Alon, 2002)) are defined as subgraph patterns recurring in real-world networks at frequencies that are statistically significant from those in random networks.…”

“…Note that if |C| > 1, then r, D, and B are from the largest component. Pržulj et al, 2004). Previous work showed that graphlets can be used to define universal classes of networks (Milo et al, 2002).…”

“…Moreover, graphlets are at the heart and foundation of many network analysis tasks (e.g., network classification, network alignment, etc.) (Pržulj et al, 2004;Milenkoviae and Pržulj, 2008;Hayes, Sun and Pržulj, 2013). In this paper, we introduce an efficient algorithm to compute the number of embedding of each graphlet of size k = {2, 3, 4} nodes in the network (see Table 1 for notation).…”