|budding yeast (S. cerevisiae)|
|average degree: 3.515|
|clustering coefficient: 0.146|
|average path length: 5.37|
|BioGRID physical and genetic interactions|
|HitPredict small-scale 'high-confidence' data, and predicted interactions based on Bayesian inference|
|UniProt comprehensive and freely accessible resource of protein sequence and functional information|
|STRING database of known and predicted protein interactions, includes direct (physical) and indirect (functional) associations|
|DPiM protein interaction map of the Drosophila melanogaster proteome|
|IntAct open source database system and analysis tools for molecular interaction data|
|MINT protein-protein interactions mined from the scientific literature by expert curators|
|DIP experimentally determined interactions between proteins, combining information from a variety of sources|
|HPRD human protein reference database|
|MIPS collection of manually curated high-quality PPI data collected from the scientific literature by expert curators|
|TAIR Arabidopsis protein-protein interaction data curated from the literature|
|MatlabBGL very fast graphs package for Matlab|
|Brain Connectivity Toolbox Matlab scripts to calculate most standard graph-theoretic quantities|
|Sergei Maslov's website Matlab scripts for degree-preserving network rewiring and degree-degree correlation maps|
|modularity calculator Matlab implementation of the Louvain algorithm to calculate network modularity|
|PINA integrated platform for protein interaction network construction, filtering, analysis, visualization and management|
|Cytoscape open source software platform for visualizing complex networks and integrating these with any type of attribute data|
|MPact common access point to interaction resources at MIPS|
|SNAP general purpose, high performance system for analysis and manipulation of large networks|
|APID Agile Protein Interaction DataAnalyzer: interactive bioinformatic web-tool that has been developed to allow exploration and analysis of main currently known information about protein-protein interactions integrated and unified in a common and comparative platform|
|Gephi interactive visualization and exploration platform for all kinds of networks and complex systems, dynamic and hierarchical graphs|
The Dill research group, at Stony Brook University's Laufer Center, has recently begun a computational study of eukaryotic protein-protein interaction (PPI) network evolution. We published our basic model layout in PLoS ONE, in 2012:
J. Peterson, S. Presse, K. Peterson, and K. Dill. Simulated evolution of protein-protein interaction networks with realistic topology. PLoS ONE 7(6): e39052, 2012.
The Matlab scripts we used to carry out our simulations are freely available on GitHub. (In addition, you will need to install the MatlabBGL package and the Louvain modularity script.) In our model, protein networks evolve by two known biological mechanisms: (1) a gene can duplicate, putting one copy under new selective pressures that allow it to establish new relationships to other proteins in the cell, and (2) a protein undergoes a mutation that causes it to develop new binding or new functional relationships with existing proteins. In addition, we allow for the possibility that once a mutated protein develops a new relationship with another protein (called the target), the mutant protein can also more readily establish relationships with other proteins in the target’s neighborhood.
The visualizations shown here are generated using sigma.js, with the underlying graph files created by Gephi. The layouts are generated using Gephi's ForceAtlas2 algorithm. The data used here are the small-scale and 'hi-confidence' datasets from HitPredict. (Clicking on an organism name gives a direct link to download the data.) For the small-scale datasets, both colors and node sizes represent degree (number of interactions). For the hi-confidence datasets, node sizes represent degree, and the colors partition the graph by modularity class ('communities' of proteins).
The code for this website is open-source, and is available at GitHub.
Welcome to interacto.me, an interactive tool for visualizing protein interaction networks!
To use this tool, first, choose your organism of interest. Then, push the corresponding button at the top left, to generate the protein-protein interaction networks for that organism. The panel below those buttons gives you the option for "small-scale" vs. "high-confidence". 'Hi-confidence' uses selected high-throughput experimental data, using the HitPredict method. 'Small-scale' excludes data obtained from high-throughput experiments, as these data can be unreliable.
To drill down deeper, you can explore one protein at a time in two different ways:
Click on a node to get more information about the particular protein it represents. You'll get a zoomed-in view, plus some basic information about the protein from the UniProt database. You can click on the protein name in the popup box to visit the UniProt page, which has much more detail about the protein, including its amino acid sequence. You can also zoom in using your mouse wheel, and drag the network around on your screen by using the mouse.
To search for a particular protein or gene of interest, type the name of the protein or gene (or the UniProt/Swissprot ID), then click 'search' in the box at the upper right. For more information about how these networks are generated, click the 'about' button in the upper right hand corner.
(To display this popup again, just click 'help' on the top bar.)