User:JurisicaLab/I2D - Interologous Interaction Database

I2D - Interologous Interaction Database

Introduction

The Interologous Interaction Database (I2D) is a web-accessible database of known and predicted mammalian and eukaryotic protein-protein interactions (PPIs). The database has been developed and is maintained by Jurisica Lab[1] at Ontario Cancer Institute, Princess Margaret Hospital, UHN[2].

I2D will continue to expand as new protein-protein interaction data becomes available.

Description

The database was originally referred to as Online Predicted Human Interaction Database (OPHID) and contained known and predicted human protein-protein interactions. The database was later expanded by creating interologous networks in multiple organisms as rat, mouse, fly, worm, and yeast. It has been built by mapping literature-based interactions from BIND, BioGrid, DIP, HPRD, MINT as well as high-throughput (HTP) data between species. Thus, until experimentally verified, these interologous interactions should be considered “predictions”. Through this database the high-quality human interactome can be transferred for instance to mouse, extending the mouse interactome by tens of thousands of protein interactions.

I2D includes data for Saccharomyces cerevisiae, Caenorhabditis elegans, Drosophila melanogaster, Mus musculus, Rattus norvegicus and Homo sapiens. It remains one of the most comprehensive sources of known and predicted eukaryotic PPIs.

The over 200,000 predicted interactions currently listed in I2D, are evaluated using protein domains, gene co-expression and gene ontology terms. I2D can be queried using single or multiple IDs. The results can be displayed as text, HTML, or visualized with our custom OpenGL-accelerated network visualization software tool called NAViGaTOR(Network Analysis, Visualization and Graphing, TORonto) available for Windows, Linux, Solaris and OSX platforms. In general the results can be visualized any other graph visualization tool that supports PSI MI XML data format, such as Cytoscape and VisAnt. I2D can also be queried visually from within NAViGaTOR. In addition, the entire database is available for download in tab-delimited text or PSI-compliant XML format.

Background

Protein-Protein Interaction networks

The network of protein–protein interactions is referred to as the interactome. The interactome is a whole set of protein interactions forming the backbone of various biological pathways. Pathways can be signaling pathways which are interaction of signaling molecules like hormones, neurotransmitters, cytokines, with their cognate receptors, metabolic pathways that are interaction of enzymes and substrates, immune system pathways like antigen-antibody interaction and many other cellular processes required for normal cell function. Complete knowledge of these pathways will help in the understanding of the normal processes in the cell, as well as how diseases such as cancer develop from mutation of individual pathway components. Thus the evolution of high-throughput (HTP) technologies in the post-genomics era has taken scientists from the characterization of single proteins to the investigation of entire interactomes.

Concept of Interologs

One of the biological techniques is to map interactomes between species using orthologs, making predictions about new interactions that have not yet been demonstrated experimentally. Orthologs are proteins in different species that tend to have similar function, and so their prediction has become an important component of comparative genomics and genome annotation. Ortholog prediction for large genome-scale datasets is typically performed using a reciprocal best-hit approach (RBH) approach. Interologs are interacting proteins from one species whose orthologous proteins from another species also interact.

Construction of Interologous networks for I2D

Using the known human interactome (literature-based interactions from BIND, BioGrid, DIP, HPRD, and MINT, plus HTP screens) as a starting point, we created interologous networks in multiple organisms. PPI networks were transferred between organisms using interologs. Briefly, interactions from organism X are inferred in organism Y if the two interacting proteins from X have orthologs in Y. We generated a database of orthologs between each of the six eukaryotic organisms fly, rat, mouse, worm, yeast and human. Orthologs were then used to map the interactome of one organism into another using the reciprocal best-hit approach (RBH) approach. Blasting was carried out on an IBM p690 mainframe using NCBI stand-alone BLAST (v.2.2.14); results were parsed using DB2 Information Integrator (v.8.1.1), and compiled in an IBM DB2 database (v.8.1.6). The external databases included into I2D are SwissProt , Entrez Gene, UniGene and International Protein Index(IPI).

Statistics

Following are the statistics as of December 2009 from I2D version 1.71

Source	Source Interactions	Predicted Interactions	Total
FLY	34,096	28,965	62,770
RAT	1,770	36,833	38,084
MOUSE	5,310	72,984	76,559
WORM	7,024	25,407	31,766
YEAST	120,702	2,804	122,326
HUMAN	60,675	34,824	92,561
Total	229,577	201,817	424,066

Up to date statistics can be found on the I2D website.

Acknowledgments

We acknowledge the hardware and software support from IBM Life Sciences through a Shared University Research Grant, and support from the National Science and Engineering Research Council, the Institute for Robotics and Intelligent Systems, Precarn Inc., Genome Canada via the Ontario Genome Institute, Ontario Research Fund, Canadian Institute of Health Research, Canada Foundation for Innovation Canada Research Chair Program, the Princess Margaret Hospital Foundation through funds raised by the Toronto Women's Fashion Show Committee and Younger Foundation. We also acknowledge our software development team, Richard Lu, Frederic Breard, David Otasek, and Uzma Khan.

References

1. Brown, K.R., and Jurisica, I. (2007) Unequal evolutionary conservation of human protein interactions in interologous networks. Genome Biology, 8(5):R95. PubMed | PDF

2. Brown, K.R., and Jurisica, I. (2005) Online Predicted Human Interaction Database. Bioinformatics, 21(9):2076-82. PubMed | PDF

Authors

Brown K.R.¹ and Jurisica I.^1,2,3

1. Jurisica Lab, IBM Life Sciences Discovery Centre, Ontario Cancer Institute, Toronto Medical Discoveries Tower, 101 College Street, Toronto, Ontario, Canada.
2. Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada 3. Department of Computer Science, University of Toronto, Toronto, Ontario, Canada

Links

Jurisica Lab Homepage
I2D Homepage

Contact

juris@ai.utoronto.ca