Protein-Protein Interaction Networks, developed by

ProViz:Protein Interaction Visualization and ExplorationFlorian Iragne (1),Macha Nikolski (1),Bertrand Mathieu (1),David Auber (1)and David Sherman(1)(1)LaBRI UMR 5800,Universit´e Bordeaux 1,351cours de la Lib´e ration,33405Talence Cedex,Franceiragne@labri.fr,macha@labri.fr,mathieu@labri.fr,auber@labri.fr,david@labri.frAbstractSummary :ProViz is a tool for the visualization ofProtein-Protein Interaction Networks,developed by the IntAct european project.It provides facilities for navigating in large graphs and exploring biologically-relevant features,and adopts emerging standards such as GO and PSI-MI.Availability :ProViz is available under the GPL and may be freely downloaded.1Contact :david.sherman@labri.frIntroductionAnalysis of protein-protein interaction (PPI)net-works requires a combination of algorithmic and vi-sualization tools,ideally integrated within a software platform that is itself integrated with access to local and distant data banks.We present a software tool called ProViz that provides highly-interactive visual-ization of large networks of interactions,integrated with the IntAct data model[5].ProViz is similar in purpose to PIMrider,[8],Osprey[2],and other visual-ization or analysis tools[11,7,6,10].Overview of ProVizGraph drawing and interactive graph exploration are active domains in computer science and many tools are available for this task.Adaptation of these tools and techniques to the specific needs of biologists exploring PPI networks is a current effort in bioinfor-matics.The challenge is to add valuable information and functions that enable the user to discover interest-ing biological relations hidden within the data.ProViz improves over existing work by providing a fast,scalable,open tool with extensive plugins,that integrates emerging standards for representing biolog-ical knowledge in a biologist-oriented interface.Intended use ProViz is designed with an under-standing of the ways that biologists prefer to work.It may be used for exploring large graphs in order to identify proteins and interactions of interest,ei-ther through keyword search or through analysis of the combinatorial structure of the network;for comparing graphs from different strains or species over ortholo-gous sets of genes;for extracting views and subgraphsfor further analysis;and for clustering related proteins and interactions (see examples in Supplementary Ma-terial).ProViz is highly interactive,providing screen updates within 50ms on standard workstations while manipulating graphs with a million elements.ProViz can be a content-type helper for interaction database query results in PSI-MI format[3].Name-based or sequence-based queries to the IntAct feder-ated database of protein-protein interactions produce networks that “link out”to ProViz for detailed study,and protein nodes and interaction edges link back in to IntAct web services.User interface The ProViz screen is intentionally uncluttered (figure 1).The right half of the screen dis-plays the current view of the current graph;the differ-ent views available are selected through the use of tabs above the window.Above this window in the tool bar are buttons for changing the layout of the current view.The mouse can be used to select elements or to move elements,and the mouse wheel can be used to zoom in or out and to pan the image.Below are buttons for cloning and for closing the view.Four tabs are avail-able on the left:Views ,for information about existing views;Node Ontology ,for selecting proteins based on GO terms;Edge Ontology ,for selecting interac-tions based on controlled vocabularies;and Proper-ties ,for viewing the complete set of properties associ-ated with a node or edge element.Views Subgraphs produced by selection,filtering,or clustering are automatically organized into views that can be manipulated independently and used to produce subsequent views.Each view has its own lay-out and zoom,and views can be used to compare dif-ferent analyses of the same interaction network.Views are organized in a tree,a quotient graph whose nodes are individual subgraphs.Layout algorithms Of the dozens of layout algo-rithms in the plugin library,three were chosen for direct use based on their capacity to highlight bio-logically pertinent information.GEM [4]is an effi-cient directed force-based graph drawing algorithm.It groups related nodes and can be used to quickly iden-tify proteins with a given role,or for visualizing pro-tein complexes.Hierarchical layout [9]reveals ances-1Sourcecode and binaries are available at bri.fr/eng/proviz.htmBioinfor m atics © Oxford University Press 2004; all rights reserved.Bioinformatics Advance Access published September 3, 2004Figure1.View of a spoke-model PPI graph for yeast.(a)Protein properties(b)Protein selection using GO terms(c)Main window showing part of current view(right)and the cluster tree(left)tral relationships between nodes and is useful when looking for cascade-type interactions or comparison to metabolic pathway data.Circular layout is a neutral choice that does not attribute any semantics to edge relations.Integrating controlled vocabularies ProViz uses GO and PSI-MI controlled vocabularies for describing proteins and ers employ these vocabu-laries when building views of interaction networks by manualfiltering or through the use of clustering plug-ins.Infigure1we see the property list for the node corresponding to yeast Rad16(nucleotide excision re-pair protein),including GO evidence,gene names,and external links.Tulip Development Platform ProViz development is based on the Tulip platform[1],designed for management and three-dimensional display of large graphs.It provides a rich set of operations on graphs: metric computation,node and edge layout,selection, extraction of view and subgraphs,and labeling of nodes and edges with arbitrary sets of attributes.Oper-ations specific to the application domain are provided by means of software plugins.Any program using Tulip can add to the core features by providing its own domain-specific plugins.Tulip is written in C++and uses Qt and OpenGL for enhanced portability. AckowledgementsThis work is supported by EU grant number QLRI-CT-2001-00015under the RDD programme“Quality of Life and Management of Living Resources.”References[1]David Auber.Tulip-a huge graph visualizationframework.In M.Juenger and P.Mutzel,editors,Graph Drawing Software,Mathematics and Vi-sualization,pages80–102,Heidelberg,August 2003.Springer-Verlag.[2]BJ.Breitkreutz,C.Stark,and M.Tyers.Osprey:a network visualization system.Genome Biol-ogy,4(3):R22,2003.[3]H.Hermjakob et al.(39authors).The HUPOPSI’s molecular interaction format–a community standard for the representation of protein interac-tion data.Nat.Biotechnol.,22(2):177–83,Feb.2004.[4]A.Frick,A.Ludwig,and H.Mehldan.A fastadaptive layout algorithm for undirected graphs.In Springer-Verlag,editor,Proc.Workshop on Graph Drawing94,volume LNCS894,pages 389–403,1994.[5]H.Hermjakob,L.Montecchi-Palazzi,C.Lew-ington,S.Mudali,S.Kerrien,S.Orchard, M.Vingron,B.Roechert,P.Roepstorff,A.Va-lencia,H.Margalit,J.Armstrong,A.Bairoch,G.Cesareni, D.Sherman,and R.Apweiler.Intact:an open source molecular interaction database.Nucleic Acids Res.,32:D452–5,Jan.2004.[6]T.Koike and A.Rzhetsky.A graphic editor foranalysing signal-transduction pathways.Gene, 259:235–244,2000.[7]ppe,J.Park,O.Niggermann,and L.Holm.Generating protein interaction maps from in-complete data:application to fold assignment.Bioinformatics,17(1):S149–S156,2001.[8]P.Legrain,J.Wojcik,and JM.Gauthier.Protein-protein interaction maps:a lead towards cellular functions.Trends in Genetics,17,2001.[9]E.B.Messinger,L.A.Rowe,and R.R.Henry.Adivide and conquer algorithm for the automaticlayout of large directed graphs.IEEE Transac-tions on Systems,Man and Cybernetics,SMC-21(1):1–11,Jan/Feb1991.[10]P.Shannon,A.Markiel,O.Ozier,N.Baliga,J.Wang,D.Ramage,N.Amin,B.Schwikowski,and T.Ideker.Cytoscape:a software environ-ment for integrated models of biomolecular in-teraction networks.Genome Res.,13(11):2498–504,Nov.2003.[11]CL.Tucker,JF.Gera,and P.Uetz.Towardsan understanding of complex protein networks.Trends in Cell Biol.,11(3):102–106,2001.。