Application Note

^*abstract

Biological processes involve complex networks of interactions between molecules. Various large-scale experiments and curation efforts have led to preliminary versions of complete cellular networks for a number of organisms. To grapple with these networks, we present developed tYNA, a Web system for managing, comparing and mining multiple networks, both directed and undirected. tYNA efficiently implements methods that have proven useful in network analysis, including locating identifying defective cliques, finding small network motifs (such as feed-forward loops), calculating global statistics (such as the clustering coefficient and eccentricity), and identifying hubs and bottlenecks. It also allows one to manage a large number of private and public networks using a flexible tagging system, to filter them based on a variety of criteria, and to visualize them through an interactive graphical interface. A number of commonly used biologically datasets have been pre-loaded into tYNA, standardized and grouped into different categories.

Availability: The tYNA system can be accessed at http://networks.gersteinlab.org/tyna. The source code, JavaDoc API, and WSDL can also be downloaded from the web site. tYNA can also be accessed from the Cytoscape software using a plugin.

Supplementary Information: Additional figures and tables can be found at http://networks.gersteinlab.org/tyna/supp.

1. introduction

In the era of systems biology, the focus on understanding complex organisms is shifting from studying individual genes and proteins towards the relationships between them (Sharan and Ideker, 2006). These relationships are usually expressed in terms of various kinds of biological networks, including protein-protein interactions, gene regulation, and metabolic pathways. Recent developments of large-scale experiments such as mass spectrometry and array-based techniques (Lee et al., 2002, Gavin et al., 2006, Krogan et al., 2006) have generated rough descriptions of the complete networks in the cells. Many studies have reported interesting biological findings from these networks, including the relationships between various statistical properties of a gene and its function and essentiality, and the elucidation of controls at the molecular level based on network motifs (Lee et al., 2002, Milo et al., 2002, Shen-Orr et al., 2002, Yu et al., 2003, Han et al., 2004, Yu et al., 2004).

 These studies require heavy computations on multiple networks. We have developed a Web system, tYNA (TopNet-like Yale Network Analyzer), to provide researchers with a set of tools to carry out such computations with great ease. The system provides five main types of functionality: (1) Network management: storing, retrieving and categorizing networks, as well as performing format conversion. A comprehensive set of widely used network datasets is preloaded, put into standard form, and categorized with a set of tags. (2) Network visualization: displaying networks in an interactive graphical interface, with interactive controls for zooming, panning, coloring and exporting networks (Fig. 1). (3) Network comparison and manipulation: various kinds of filtering based on network statistics and node names, and multiple network operations such as taking the union and intersection of networks.  (4) Network analysis: computing various statistics for the whole network and subsets, and finding motifs and defective cliques. (5) Network Mining: predicting one network based on the information in another.

Fig. 1.The intersection of two yeast-two-hybrid datasets (Ito et al., 2000, Uetz et al., 2000) with all nodes having no edges in the intersection filtered by a statistics filter. The nodes are colored according to their degrees. Also shown in the figure are the various statistics of the resulting network.

Our system shares some elements with a number ofsome other recently developed network analysis and visualization systems, includingsuch as BioGrid¹, Cytoscape (Shannon et al., 2003), JUNG²¹, N-Browse³², and Osprey⁴³, but also offers some unique additional features such as defective clique finding. Besides, being a Web-based system, tYNA also has some unique advantages:

• Users can share networks through a centralized database.
• Computationally intensive tasks such as motif finding and statistics calculations can be performed on powerful servers.
• The system can be linked from/to other online resources.
• Users can incorporate some functions of tYNA into their own programs using the SOAP-based web service interface.

Table 1 summarizes some major differences between the systems. A more detailed description is given on the supplementary web site.

Table 1.A brief comparison of several existing systems

2. Using tYNA

tYNA provides a simple view with some basic features, and an advanced view for more complex analyses. One may start with the simple view, and switch to the advanced view later after getting familiar with the interface.

1. Fig. 2 shows the typical workflow of using tYNA. In the following we explain the various components in some detail.Fig. 2.Uploading networks and categories

Login is optional. However, some features such as the management of private networks are for registered users only. The first step of analysis is to upload networks. tYNA accepts various file formats, including the SIF format of Cytoscape, and it can handle both directed and undirected networks, as well as simple and multigraph networks. One may also enter additional attributes to organize the networks into groups, such as network type (e.g. protein-protein interaction), organism (e.g. yeast) and experimental method (e.g. yeast-two-hybrid) (Fig. S1). Furthermore, tYNA allows users to analyze subsets of the networks (e.g., active parts in a dynamic network (Han et al., 2004, Luscombe et al., 2004)) by using category files.. For that purpose, category files grouping the nodes into different categories can be uploaded.

2. Loading networks into workspaces

After uploading a network, one may view its statistics and visualize it graphically by loading it into a workspace. A workspace is a working area for a single network (Fig. 1). There are several workspaces, which allow one to work on multiple networks at the same time. Various statistics are computed for each node, such as the degree, clustering coefficient, eccentricity and betweenness (Yu et al., 2004). The average, minimum, maximum and standard deviation of each statistic in the network and each category-defined sub-network are displayed, as well as the number of nodes, edges, and connected components. Networks are visualized in Scalable Vector Graphics (SVG) using the aiSee package⁵package⁴, which facilitates an interactive interface: one may change the appearance of the network such as zoom level and coloring in real time (Fig. S2, S3).

3. Single-network operations (advanced view)

After loading a network into a workspace, one may perform filtering and motif finding. Filtering allows one to retain a portion of the network, based on a statistics cutoff (e.g. the 5% of nodes with the highest out-degrees) or a regular expression relating to node names. They It will easily allow one to identify the hubs and bottlenecks in a graph. Motif finding identifies various regular patterns in the network, including chains, cycles, feed-forward loops and complete two-layers. They generalize the motifs discussed in previous studies (Lee et al., 2002, Milo et al., 2002, Shen-Orr et al., 2002). Currently all occurrences of a specified motif pattern are returned. We will study the feasibility of returning only statistically over-represented motifs in future work. tYNA also identifies defective cliques (Yu et al., 2006) that suggest potential missing edges in a network (Fig. S4, S5).

4. Multiple-network operations (advanced view)

Multiple-network operations allow one to select multiple networks, perform some operations on each of them, and merge them into a single network. For example, the intersection of multiple high-throughput protein-protein interaction networks offers a high-confidence set of potential interactions. The relationships between different kinds of networks, such as gene regulation and co-expression, can also be studied (Fig. S6).

5. Mining and Edge overlap (advanced view)

The edge overlap feature allows the comparison of the edges in two networks. It can test, using some prediction functions, how well one network predicts another. Some prediction functions are predefined, such as identity, sibling and couple (Table S1S2).

6. Saving and downloading analyzed networks

Finally, one may save a working network into the database, or send it to another workspace (as a temporary backup). Likewise, one may download a working network in various network and graphics formats, including SIF, bitmap, postscript and PDF (Fig. S3, S4).

3. Implementation

All source codes were written in Java using standard J2EE architecture. A detailed JavaDoc API is available for users who want to use the classes in their own codes. The tYNA database can also be accessed through standard SOAP-based web services, and we have developed a plugin (available on the tYNA website) to interface tYNA with the network visualization system Cytoscape (Fig. S7).

4. Discussion

As biological network analysis is a vigorous research area, new statistics, motifs, and mining algorithms are expected to emerge continuously. tYNA was thus designed in a modular fashion so that new features can be readily added. Being a Web system, the new features are immediately made accessible to users without requiring them to perform software updates.

 We plan on connecting tYNA with dataset management systems such as YeastHub (Cheung et al., 2005), and Bind (Alfarano et al., 2005), DIP (Xenarios et al., 2002), MINT (Zanzoni et al., 2002), Reactome (Joshi-Tope et al., 2005) and annotation databases to provide a unified platform for performing complex analyses. We think that the combination of the analysis features provided by tYNA and the advanced visualization facilities of Cytoscape can prove particularly powerful. We also plan on interfacing tYNA with analysis and visualization tools such as bioPIXIE (Myers et al., 2005) and Pajek (Batagelj and Mrvar, 2003), which would allow researchers to combine the distinct features of each tool.

References

Alfarano C. et al. (2005) The Biomolecular Interaction Network Database and Related Tools 2005 Update. Nucleic Acids Research 33 D418-D424.

Batagelj V. and Mrvar A. (2003) Pajek – Analysis and Visualization of Large Networks. Jünger, M., Mutzel, P., (Eds). Graph Drawing Software, Springer, 77-103.

Cheung K. H. et al. (2005) YeastHub: A Semantic Web Use Case for Integrating Data in the Life Sciences Domain. Bioinformatics 21 Suppl1 i85-i96.

Gavin A-C. et al. (2006) Proteome Survey Reveals Modularity of the Yeast Cell Machinery. Nature 440 631-636.

Han J-D H. et al. (2004) Evidence for Dynamically Organized Modularity in the Yeast Protein-protein Interaction Network. Nature 430 88-93.

Ito T. et al. (2000) Toward a Protein-Protein Interaction Map of the Budding Yeast: A Comprehensive System to Examine Two-hybrid Interactions in all Possible Combinations between the Yeast Proteins. Proc. Natl. Acad. Sci. USA 98 4569-4574.

Joshi-Tope G. et al. (2005) Reactome: A Knowledgebase of Biological Pathways. Nucleic Acids Research 33 D428-D432.

Krogan N. J. et al. (2006) Global Landscape of Protein Complexes in the Yeast Saccharomyces cerevisiae. Nature advance online publication.

Lee T. I. et al. (2002) Transcriptional Regulatory Networks in Saccharomyces cerevisiae. Science 298 799-804.

Luscombe N. M. et al. (2004) Genomic Analysis of Regulatory Network Dynamics Reveals Large Topological Changes. Nature 431 308-312.

Milo R. et al. (2002) Network Motifs: Simple Building Blocks of Complex Networks. Science 298 824-827.

Myers et al. (2005) Discovery of Biological Networks from Diverse Functional Genomic Data. Genome Biology 6, R114.

Shannon P. et al. (2003) Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks. Genome Research 13 2498-2504.

Sharan R. and Ideker T. (2006) Modeling Cellular Machinery through Biological Network Comparison. Nature Biotechnology 24 427-433.

Shen-Orr S. S. et al. (2002) Network Motifs in the Transcriptional Regulation Network of Escherichia coli. Nature Genetics 31 64-68.

Uetz P. et al. (2000) A Comprehensive Analysis of Protein-protein Interactions in Saccharomyces cerevisiae. Nature 403 623-627.

Xenarios I. et al. (2002) DIP, the Database of Interacting Proteins: A Research Tool for Studying Cellular Networks of Protein Interactions. Nucleic Acids Research 30 303-305.

Yu H. et al. (2004) TopNet: A Tool for Comparing Biological Subnetworks, Correlating Protein Properties with Topological Statistics. Nucleic Acids Research 32 328-337.

Yu H. et al. (2006) Predicting Interactions in Protein Networks by Completing Defective Cliques. Bioinformatics 22 823-829.

Zanzoni A. et al. (2002) MINT: A Molecular INTeraction Database. FEBS Letters 513 135-140.