Manual - Bioinformatics and Systems Biology
Contents
1. x Using the method signature as shown here the layout algorithm can use data of a node attribute to calculate the layout af public TaskIterator createTaskIterator CyNetworkView networkView Object context Set lt View lt CyNode gt gt nodesToLayOut String attrName return new TaskIterator new SomeLayoutTask toString networkView nodesToLayOut SomeLayoutContext context attrName which node attribute should be used to perform the grouping on undoSupport Override public Set lt Class lt gt gt getSupportedNodeAttributeTypes Set lt Class lt gt gt ret new HashSet lt Class lt gt gt ret add Integer class ret add Double class ret add String class ret add Boolean class add other classes if the layout can support it return ret Override public SomeLayoutContext createLayoutContext return new SomeLayoutContext Jr Override A 3 GUIDELINES CONCERNING CYTOSCAPE PLUGIN DEVELOPMENT 66 67 68 69 70 71 72 ANa kWNEH PER PR RWWWWWWWWWWNNYNNNNNNNNNKP RP RPE RP RP RRP Re FWNrFOKOANDOTKBRWNFODOOANDOBWNFOCOUANDOTBWNrF OCHO Plant Systems Biology department VIB UGent 17 public boolean getSupportsSelectedOnly return false if the layout algorithm can not work on a Is set ot selected nodes but amily Om Jul modes Omes return true J The class above does not perform the layout but is used by Cytoscape to get an instance of a2. 6 Neg cutoff 0 6 fa Pos cutoff 0 8 f To clear all input click Reset form O Save output 11 For more information about 7 ame Mew which gene ID s can be used click 8 Run analysis ID s 12 Figure B 2 Overview of a family wise coexpression graph This graph was created by using 11 bait genes from four different species along with four gene expression datasets one for each species The positive cutoff was set to 0 8 and the negative cutoff was set to 0 6 APPENDIX B COEXPNETVIZ USER MANUAL 28 Comparative Genomics Co expression Networks Figure B 3 Detail of the gene families The upper right node bright green is a family containing genes that are coex pressed with 7 out ofthe 11 bait genes and is likely involved in the same bio logical process as the bait genes Gene names of the coexpressed genes are dis played on top of each node Figure B 4 Detail of the bait genes A close up of the Malus domesticus ap ple bait genes is shown Add Edit Select Group Nested Networks A N dy M e N i ij Plaza CoExpNetviz External Links vvbazavvvyvy Extend Network by public interaction database Monocots ORTHOO3M006664 Plaza Dicots ORTHOO3D008792 Preferences a Using the context menu b Linkouts to PLAZA families Figure B 5 Linking gene families to the PLAZA website B 5 FILE FORMATS Plant Syst
3. Proost et al 2015 viii 3 5 9 12 14 PLAZA family A family of putative homologous protein encoding genes computed by PLAZA Proost et al 2015 To compute gene families the Tribe MCL algorithm is used Enright et al 2002 followed by a post processing algorithm to identify outliers In short an all vs all BLASTp Altschul et al 1990 of all genes in the PLAZA database is performed to calculate sequence similarity scores for every possible gene pair The sequence similarity scores are then normalized and used as a distance measure to perform Markov clustering van Dongen 2000 this is the Tribe MCL algorithm as described by Enright et al 2002 Finally the post processing algorithm is applied to remove outliers The post processing algorithm removes a gene from a gene family if it shows sequence similarity to only a limited number of genes in this family 9 RBH Reciprocal best hit 14 SCC Spearman s rank correlation coefficient 10 11 target gene Any gene that is present in the microarray data which was used in CoExp NetViz that is not a bait gene 7 9 TF transcription factor 11 vill Chapter 1 Introduction 1 1 Function prediction via coexpression analysis It is known that genes participating in the production of a certain metabolite tend to have more similar expression patterns than expected by chance Thus in order to discover how a plant produces a certain metabolite one could use transcriptome analysis
4. doi 10 1093 pep peu050 URL http www ncbi nlm nih gov pubmed 24683037 Movahedi S Van Bel M Heyndrickx K S and Vandepoele K Comparative co expression analysis in plant biology Plant cell amp environment 35 10 1787 98 Oc tober 2012 ISSN 1365 3040 doi 10 1111 j 1365 3040 2012 02517 x URL http www ncbi nlm nih gov pubmed 22489681 Mutwil M Obro J Willats W G T and Persson S GeneCAT novel webtools that combine BLAST and co expression analyses Nucleic acids research 36 Web Server issue W320 6 July 2008 ISSN 1362 4962 doi 10 1093 nar gkn292 URL http www pubmedcentral nih gov articlerender fcgi artid 2447783ktool pmcentrez amp rendertype abstract Mutwil M Klie S Tohge T Giorgi F M Wilkins O et al PlaNet com bined sequence and expression comparisons across plant networks derived from seven species The Plant cell 23 3 895 910 March 2011 ISSN 1532 298X doi 10 1105 tpc 111 083667 URL http www pubmedcentral nih gov articlerender fcgi artid 3082271 amp tool pmcentrez amp rendertype abstract Proost S Van Bel M Vaneechoutte D Van de Peer Y Inze D et al PLAZA 3 0 an access point for plant comparative genomics Nucleic acids research 43 Database issue D974 81 January 2015 ISSN 1362 4962 doi 10 1093 nar gku986 URL http nar oxfordjournals org content 43 D1 D974 BIBLIOGRAPHY Plant Systems Biology department VIB UGent 33 Rasmussen S Barah
5. to lt Embed Dependency gt scope provided test lt Embed Dependency gt As this will prevent the test dependencies to be embedded in the plugin jar file Writing a JUnit test It is recommended to follow the default Maven folder structure when creating JUnit tests For example if the class to test is found at PLUGIN_ROOT src main java com something package_name ClassName java then the test class should be found at PLUGIN_ROOT src test java com something package_name ClassNameTest java Also any resources used by test classes should be found in this folder or a subdirectory of this folder PLUGIN_ROOT src test resources The test class is then located in the package_name package in the test folder it should have the following minimal layout package com something package_name import org junit After import org junit AfterClass import org junit Before import org junit BeforeClass import org junit Ignore import org junit Test import static org junit Assert add other imports if needed public class ClassNameTest public ClassNameTest implement methods setUpClass tearDownClass setUp and tearDown if necessary Test of firstMethod method of class className Test public void testFirstMethod add assertions here tests of other methods here APPENDIX A DEVELOPMENT OF COEXPNETVIZ HovosooutkumnmrH Rp OANnDwWIBWNH m CONDOKRWNH 14 C
6. J et al The Reactome pathway knowledgebase Nucleic acids research 42 Database issue D472 7 January 2014 ISSN 1362 4962 doi 10 1093 nar gkt1102 URL http www pubmedcentral nih gov articlerender fcgi artid 3965010 amp tool pmcentrez amp rendertype abstract Daub C O Steuer R Selbig J and Kloska S Estimating mutual information us ing B spline functions an improved similarity measure for analysing gene expression data BMC bioinformatics 5 118 August 2004 ISSN 1471 2105 doi 10 1186 1471 2105 5 118 URL http www pubmedcentral nih gov articlerender fcgi artid 516800 amp tool pmcentrez amp rendertype abstract De Bodt S Hollunder J Nelissen H Meulemeester N and Inze D CORNET 2 0 integrating plant coexpression protein protein interactions regulatory interac tions gene associations and functional annotations The New phytologist 195 3 707 20 August 2012 ISSN 1469 8137 doi 10 1111 j 1469 8137 2012 04184 x URL http www ncbi nlm nih gov pubmed 22651224 Enright A J Van Dongen S and Ouzounis C A An efficient algorithm for large scale detection of protein families Nucleic acids research 30 7 1575 84 April 2002 ISSN 1362 4962 URL http www pubmedcentral nih gov articlerender fcgi artid 101833 amp tool pmcentrez amp rendertype abstract 31 32 Comparative Genomics Co expression Networks Hansen B O Vaid N Musialak Lange M Janowski M and Mutwil M Elucidating gene funct
7. SomeLayoutTask which does the actual work The second class that must be created is this task package com something package_name import org cytoscape model CyNetwork import org cytoscape model CyNode import org cytoscape model CyTable import org cytoscape view layout AbstractLayoutTask import org cytoscape view model CyNetworkView import org cytoscape view model View import org cytoscape view presentation property BasicVisualLexicon import org cytoscape work TaskMonitor import org cytoscape work undo UndoSupport public class SomeLayoutTask extends AbstractLayoutTask private TaskMonitor taskMonitor private CyNetwork network private SomeLayoutContext context public SomeLayoutTask final String displayName CyNetworkView networkView Set lt View lt CyNode gt gt nodesToLayOut SomeLayoutContext context String attrName UndoSupport undo super displayName networkView nodesToLayOut attrName undo this context context J ERER This method is called by Cytoscape to perform the layout Override final protected void doLayout final TaskMonitor taskMonitor here starts the actual work of placing the nodes to get the node instances do for View lt CyNode gt nv nodesToLayOut 4 CyNode node nv getModel node is now a CyNode instance to place it at a certain coordinate do networkView getNodeView node setVisualProperty BasicVisualLexicon NODE_X_LOCATION x
8. e The approach explained here will only work for libraries that are available in the Cy toscape Maven repositories or the central Maven repositories http mvnrepository com e Even if a certain third party library is already embedded in the Cytoscape core program as is the case for httpcomponents it is still recommended to embed it again in a plugin Otherwise a new Cytoscape release which uses different versions of these third party libraries can cause the plugin to break A 3 3 Adding JUnit tests To use Junit tests two extra dependencies junit and mockito are required see also ap pendix A 3 2 As these dependencies are only required during testing they should be given the test scope This can be done as follows lt dependency gt lt groupld gt junit lt groupld gt lt artifactId gt junit lt artifactId gt lt version gt 4 10 lt version gt lt scope gt test lt scope gt lt dependency gt lt dependency gt lt groupld gt org mockito lt groupld gt lt artifactId gt mockito all lt artifactId gt lt version gt 1 10 19 lt version gt lt scope gt test lt scope gt lt dependency gt Also the following line should be changed in the felix setup in the pom xm1 lt Embed Dependency gt scope provided lt Embed Dependency gt A 3 GUIDELINES CONCERNING CYTOSCAPE PLUGIN DEVELOPMENT ONaAoKRWN m NNNNN NNNNNNA RI RRP EEL SAISON HS SAU RUN Ho Plant Systems Biology department VIB UGent 13
9. expression values not probe intensities Also if you which to use log transformed APPENDIX B COEXPNETVIZ USER MANUAL 26 Comparative Genomics Co expression Networks data you must first do the log transform yourself as CoExpNetViz will not transform or alter the data in any way Finally avoid having many similar conditions in the microarray dataset this will lead to columns with redundant information and many very high PCCs Gene family files The gene family files must follow the following format every line starts with the name of a gene family followed by a tab character and a list of genes belonging to that family the genes are also tab separated The file may not contain a column header B 5 FILE FORMATS Plant Systems Biology department VIB UGent 27 Settings Load Save Delete 9 10 Reset form Title optional 8 Input bait genes Upload file with bait genes Enter bait genes 118 LO p s a Choose file with bait genes 3 Browse Choose which data sets to use pick one dataset for every species you have specified in the bait genes 42 E Figure B 1 CoExpNetViz input Ab Een form Fill in the form and click Run analysis 8 to start the coexpression analysis To save a specific configu ration click Save then click Load to load it again later or click Delete to remove the saved configuration 10 Ei 5 Add species Choose cutoff
10. the CoExpNetViz user manual which is also available at the CoExpNetViz web site See also the Github repository https github com SamDM CoExpNetViz and ap pendix A for development information 23 24 Comparative Genomics Co expression Networks B 1 Installation The normal way to install a Cytoscape plugin is with the App Manager but as CoExp NetViz is not yet published this installation method is unavailable To install the plugin download the program from the CoExpNetViz website at http bioinformatics psb ugent be webtools coexpr index php and copy it to the folder CYTOSCAPE_HOME CytoscapeConfiguration 3 apps installed In most cases the CYTOSCAPE_HOME directory which is the directory where Cytoscape is installed is located in the user home folder If Cytoscape is already running the app can be used immediately if not then the next time Cytoscape is started the app will be installed Important To run the app Cytoscape version 3 1 or higher should be used It is strongly recommended to use java 1 6 or 1 7 to run Cytoscape Java 1 8 might also appear to work fine but can cause strange behavior in Cytoscape B 2 Usage of CoExpNet Viz To start the app go to Apps gt CoExpNetViz in the Cytoscape menu clicking this entry will launch a form where gene expression datasets can be submitted To find out which genes are coexpressed with your genes of interest enter these genes in the bait genes text field 2 in f
11. De Bodt et al 2012 Table 1 1 Comparison of community accessible coexpression based gene discovery tools for plants 1 According to Mutwil et al 2011 the PlaNet algorithm can be down loaded and tested locally with any microarray data But currently as of April 2015 the PlaNet downloads page http aranet mpimp golm mpg de download htm1 does not allow downloading the program 2 Highest Reciprocal Rank see http aranet mpimp golm mpg de faq html 3 Pearson Correlation Coefficient tract densely connected subgraphs which often contain genes belonging to same biological process Networks of individual genes as described above can become very large and complex To resolve this groups of homologous genes can be represented as one single node and edges can be filtered out if they are present in only one or a few species This results in a family network with only conserved coexpression links where nodes represent families of homologous genes Interpretation of such networks and interpretation of network topology is then less complicated CoExpNetViz CoExpNetViz is available as a Cytoscape plugin Shannon et al 2003 Saito et al 2012 and as a web tool After choosing bait genes and microarray datasets in the Cytoscape plugin the analysis is run and the resulting network is displayed imme diately Using the web tool the user can download the Cytoscape files and import them manually into the program Additionally the use
12. L http www pubmedcentral nih gov articlerender fcgi artid 403769 amp tool pmcentrez amp rendertype abstract Steinhauser D Usadel B Luedemann A Thimm O and Kopka J CSB DB a comprehensive systems biology database Bioinformatics 20 18 3647 3651 July 2004 ISSN 1367 4803 doi 10 1093 bioinformatics bth398 URL http www ncbi nlm nih gov pubmed 15247097 Steuer R Kurths J Daub C O Weise J and Selbig J The mutual informa tion detecting and evaluating dependencies between variables Bioinformatics Ox ford England 18 Suppl 2 5231 40 January 2002 ISSN 1367 4803 URL http www ncbi nlm nih gov pubmed 12386007 Stuart J M Segal E Koller D and Kim S K A gene coexpression network for global discovery of conserved genetic modules Science New York N Y 302 5643 249 55 October 2003 ISSN 1095 9203 doi 10 1126 science 1087447 URL http www ncbi nlm nih gov pubmed 12934013 The Arabidopsis Genome Initiative Analysis of the genome sequence of the flowering plant Arabidopsis thaliana Nature 408 6814 796 815 December 2000 ISSN 0028 0836 doi 10 1038 35048692 URL http dx doi org 10 1038 35048692 The Potato Genome Sequencing Consortium Genome sequence and analysis of the tuber crop potato Nature 475 7355 189 95 July 2011 ISSN 1476 4687 doi 10 1038 naturel0158 URL http dx doi org 10 1038 nature10158 BIBLIOGRAPHY 34 Comparative Genomics Co expression Networks Th
13. P Suarez Rodriguez M C Bressendorff S Friis P et al Transcriptome responses to combinations of stresses in Arabidopsis Plant phys iology 161 4 1783 94 April 2013 ISSN 1532 2548 doi 10 1104 pp 112 210773 URL http www pubmedcentral nih gov articlerender fcgi artid 3613455 amp tool pmcentrez amp rendertype abstract Rhee S Y and Mutwil M Towards revealing the functions of all genes in plants Trends in plant science 19 4 212 21 April 2014 ISSN 1878 4372 doi 10 1016 j tplants 2013 10 006 URL http www sciencedirect com science article pii S1360138513002343 Saito R Smoot M E Ono K Ruscheinski J Wang P L et al A travel guide to Cy toscape plugins Nature methods 9 11 1069 76 November 2012 ISSN 1548 7105 doi 10 1038 nmeth 2212 URL http www pubmedcentral nih gov articlerender fcgi artid 3649846 amp tool pmcentrezkrendertype abstract Sewelam N Oshima Y Mitsuda N and Ohme Takagi M A step towards under standing plant responses to multiple environmental stresses a genome wide study Plant cell amp environment 37 9 2024 35 September 2014 ISSN 1365 3040 doi 10 1111 pce 12274 URL http www ncbi nlm nih gov pubmed 24417440 Shannon P Markiel A Ozier O Baliga N S Wang J T et al Cytoscape a software environment for integrated models of biomolecular interaction networks Genome research 13 11 2498 504 November 2003 ISSN 1088 9051 doi 10 1101 gr 1239303 UR
14. PI is not perfect and sometimes bugs in the plugin are happening due to bugs in the Cytoscape API In that case it can help to take a look into the Cytoscape source code to see where the problem arose There is no better way to explore the source code of a large program such as Cytoscape then by compiling it yourself and running it with a graphical debugger attached Adequate instructions to compile Cytoscape from source are available on the Cy toscape GitHub page https github com cytoscape cytoscape Many dependen cies will be downloaded when compiling Cytoscape for the first time which can result in a compile time of up to two hours After the first compilation when all dependencies are already met the compilation will take between five and fifteen minutes depending on the hardware of the computer The Cytoscape Maven parent project can be imported into NetBeans in the same way as explained in appendix A 4 Once the Cytoscape project is imported into NetBeans and compiled either from the command line or from within NetBeans it can be started in debug mode in the same way as explained in appendix A 4 1 Everything the graph ical debugger has to offer such as placing breakpoints watching variables etc is now also possible for the Cytoscape source code this can be very helpful to solve bugs that originated from within the Cytoscape API APPENDIX A DEVELOPMENT OF COEXPNETVIZ Appendix B CoExpNetViz user manual Foreword This is
15. can be used as well To specify other gene families go to the Gene family options tab 9 in fig B 1 there you can choose up to five gene family files to be used the interface for this is very similar to 4 and 5 in fig B 1 B 1 INSTALLATION Plant Systems Biology department VIB UGent 25 CoExpNetViz will merge gene families from the same file or across submitted gene family files if they contain the same gene Also CoExpNetViz already has the PLAZA monocot and PLAZA dicot families in memory by default So merging will occur with these families as well if the submitted gene family files contain ID s that are also present in PLAZA B 3 The family wise coexpression graph If the analysis run succesfully a new graph can be seen in the Cytoscape main window the family wise coexpression graph fig B 2 In the default layout bait genes are placed at the corners of the graph as big white diamond shaped nodes The nodes in the middle are gene families and the links between gene families and baits are coexpression relationships the PCC of at least one gene in this family to the bait is greater than the given threshold Positive correlations are shown in blue while negative correlations are shown in red Additionally homology relationships between bait genes are shown as light yellow dotted lines In the default layout target genes are grouped into partitions where every partition has its own color Nodes in the same partition
16. code install Git and issue the following command in the terminal 1 git clone https github com SamDM CoExpNetViz git This will download the source code into a child directory of the directory from where the command was used The newly created directory is called CoExpNetViz and contains one child directory also called CoExpNetViz along with one hidden directory the git directory It also contains a README file and a hidden file gitignore there are many Git tutorials online that explain what these files and directories are for After downloading the source code for the plugin move two directories down in the newly created folder there you will find the pom xm1 file this a file that tells Maven how the plugin should be compiled and what dependencies are required to compile it In order to compile the plugin move to the directory containing the pom xml file and run mvn clean install as shown below make sure you have working internet connection before trying this 1 cd CoExpNetViz CoExpNetViz 2 mvn clean install A 1 COMPILING COEXPNETVIZ NOV PWMNH N Plant Systems Biology department VIB U Gent 7 Maven will now automatically download all the dependencies that are required to compile the plugin then it will automatically compile the plugin after which it will automatically run JUnit tests When running this command for the first time it might take up to a minute to compile the program Compiling it again lat
17. e Tomato Genome Consortium The tomato genome sequence provides insights into fleshy fruit evolution Nature 485 7400 635 41 May 2012 ISSN 1476 4687 doi 10 1038 nature11119 URL http dx doi org 10 1038 nature11119 Toufighi K Brady S M Austin R Ly E and Provart N J The Botany Array Resource e Northerns Expression Angling and promoter analyses The Plant journal for cell and molecular biology 43 1 153 63 July 2005 ISSN 0960 7412 doi 10 1111 j 1865 313X 2005 02437 x URL http www ncbi nlm nih gov pubmed 15960624 Usadel B Obayashi T Mutwil M Giorgi F M Bassel G W et al Co expression tools for plant biology opportunities for hypothesis generation and caveats Plant cell amp environment 32 12 1633 51 December 2009a ISSN 1365 3040 doi 10 1111 3 1365 3040 2009 02040 x URL http www ncbi nlm nih gov pubmed 19712066 Usadel B Poree F Nagel A Lohse M Czedik Eysenberg A et al A guide to using MapMan to visualize and compare Omics data in plants a case study in the crop species Maize Plant cell amp environment 32 9 1211 29 September 2009b ISSN 1365 3040 doi 10 1111 j 1365 3040 2009 01978 x URL http www ncbi nlm nih gov pubmed 19389052 van Dongen S Graph Clustering by Flow Simulation May 2000 URL http www narcis nl publication RecordID oai 3Acwi n1 3A18026 Velasco R Zharkikh A Affourtit J Dhingra A Cestaro A et al The genome of the domest
18. een the plugin requesting an analysis and a user accessing the web tool A 2 2 The GUI classes When the CoExpNetViz entry in the Apps menu is clicked an instance of GuiManager in the package be samey gui is created This class is the central control point for the GUI and has references to model classes which keep track of the GUI state When the GuiManager is initialized the settings are read and the GUI is created and shown There are three packages for the gui Firstly there is be samey gui which has the GuiManager and two classes that build swing components that together form the GUI Secondly the package be samey gui model has a two model classes that keep track of the GUI state Thirdly the be samey gui controller package has classes that control button actions Adding a new GUI element To add a new element to the GUI for example to the InpPn1 add the swing components a JButton JLabel etc in the InpPn1 constructor Then create a controller class in the be samey gui controller package which has the code to execute when an action on the new GUI element is performed The controller class should extend AbstrController java Finally set the controller class as a listener to the new GUI element in the initGui method of the GuiManager A 3 Guidelines concerning Cytoscape plugin devel opment Many of the following sections are about setting up the pom xml file and more specifically setting the OSGi options For a detailed expla
19. ems Biology department VIB UGent 29 B 6 How to submit bugs If you encountered a bug you can submit it at the CoExpNetViz Github page at https github com CoExpNetViz or you can send an email to contact person placeholder org APPENDIX B COEXPNETVIZ USER MANUAL Bibliography Altschul S F Gish W Miller W Myers E W and Lipman D J Basic local alignment search tool Journal of molecular biology 215 3 403 10 October 1990 ISSN 0022 2836 doi 10 1016 S0022 2836 05 80360 2 URL http www ncbi nlm nih gov pubmed 2231712 Ashburner M Ball C A Blake J A Botstein D Butler H et al Gene ontology tool for the unification of biology The Gene Ontology Consortium Nature genetics 25 1 25 9 May 2000 ISSN 1061 4036 doi 10 1038 75556 URL http dx doi org 10 1038 75556 Barrett T Wilhite S E Ledoux P Evangelista C Kim I F et al NCBI GEO archive for functional genomics data sets update Nucleic acids research 41 Database issue D991 5 January 2013 ISSN 1362 4962 doi 10 1093 nar gks1193 URL http nar oxfordjournals org content 41 D1 D991 full Bergmann S Ihmels J and Barkai N Similarities and differences in genome wide expression data of six organisms PLoS biology 2 1 E9 January 2004 ISSN 1545 7885 doi 10 1371 journal pbio 0020009 URL http journals plos org plosbiology article id 10 1371 journal pbio 0020009 Croft D Mundo A F Haw R Milacic M Weiser
20. er should not take more than ten seconds If the plugin was compiled successfully something that looks like the following can be seen in the terminal INFO Total time 9 677 s INFO Finished at 2015 04 19T16 24 01 02 00 INF0 Final Memory 32M 221M A 1 3 Running and testing CoExpNetViz If the build was successful the compiled plugin called CoExpNetViz 1 0 SNAPSHOT jar can be found in the target folder To test the plugin copy it to your Cytoscape apps folder This folder is usually located in a subdirectory of the CytoscapeConfiguration folder which can be found in the location were Cytoscape was installed The full path is replace CYTOSCAPE_HOME by the cytoscape intallation folder CYTOSCAPE_HOME CytoscapeConfiguration 3 apps installed Once the jar file is copied to the mentioned folder launch Cytoscape if everything went well the Apps menu should now contain an entry called CoExpNetViz click the entry to launch the plugin To avoid having to copy the jar every time you make a change to the source code you can when using a Unix system symlink the jar in the target folder to the Cytoscape apps installed folder Whenever the plugin is recompiled Cytoscape will detect that the symlink in the apps folder is updated and it will reload the newly compiled plugin automatically It is recommended to run Cytoscape from the command line this gives you access to the command line interface of Cytoscape where you can manual
21. ervice api lt artifactld gt lt version gt 3 1 0 lt version gt lt scope gt provided lt scope gt lt dependency gt The Embed Transitive tag tells OSGi to embed transitive dependencies dependencies of your dependencies as well Finally the Import Package tag tells OSGi to embed all pack ages specified in the pom xm1 because of but the resolution optional will prevent embedding a package which has optional set to true if it is never imported by any other package This way all the packages of dependencies which have optional set to true will APPENDIX A DEVELOPMENT OF COEXPNETVIZ outrumx m oo n outkuwm FR 12 Comparative Genomics Co expression Networks not be embedded in the plugin jar file if they are not needed To make a dependency optional add the line lt optional gt true lt optional gt as in the example below lt dependency gt lt groupld gt org apache httpcomponents lt grouplId gt lt artifactId gt httpclient lt artifactId gt lt version gt 4 4 lt version gt lt optional gt true lt optional gt lt dependency gt Conclusion To add a third party library 1 add the three lines mentioned at the beginning of this section 2 set the scope of all Cytoscape API dependencies to provided and 3 make all third party dependencies optional After doing this the packages can be imported from within the plugin code and their classes will be available for use from within the plugin code Notes
22. g Therefore if the plugin cannot be activated looking at this file can show the cause of the problem For some good advice on an efficient development workflow see also appendix A 4 A 2 General outline CoExpNetViz is being developed in two parts one part is the Cytoscape interface which is discussed in this document another part is the web server The Cytoscape interface provides the user with a form where gene expression datasets can be chosen and param eters for the coexpression analysis can be specified The settings and files are then sent to the server which will run the actual algorithm The response is sent back to the Cy toscape plugin which will read the network files and node attribute files convert them into a network and apply the layout algorithm and visual style There are two reasons why the application is split into a local Cytoscape plugin and a web server The first reason is that the idea of making a Cytoscape plugin arose after development of the core algorithm had been started in C Cytoscape is programmed in java which is a cross platform language and the Cytoscape app installer does not support platform specific downloads This means in order to make a plugin that can run locally compiled versions of the C core for all platforms should be embedded in the plugin jar which would make the jar extremely big Another solution would be to let the user manually download the platform specific C core and somehow i
23. h can be used to visualize this data In such a network nodes represent genes while edges represent homology or coexpression rela tionships One to one one to many and many to many relationships become clear when using a network representation Also many types of information can be visualized on a network For example edge color can be used to show the type of relationship while edge width can represent the degree of similarity between coexpressed genes and node color or shape can represent the species Additionally Gene Ontology GO annotations Ashburner et al 2000 KEGG or Reactome pathway information Kanchisa et al 2014 Croft et al 2014 or MapMan functional categories Usadel et al 2009b can be mapped onto the network to help the interpretation Also network statistics can be applied to ex 2 Comparative Genomics Co expression Networks Tool CoExpNetViz PlaNet GeneCat CORNET Comparative Yes Yes Yes Yes Number of supported upto5 7 l or 2 l or 2 species Own data Yes No No Yes User interface Web interface and Web interface Web interface Web interface Cytoscape plugin Visualization Cytoscape Web page and file Web page Cytoscape import to Cytoscape is possible Output files Cytoscape network Cytoscape network None Cytoscape network and attribute files and attribute files and attribute files SVG images Similarity metric Pcc HRR PCC PCC Reference Unpublished Mutwil et al 2011 Mutwil et al 2008
24. have links to same set of baits To get more information about a node right click the node and choose Apps CoExpNet Viz fig B 5a A small dialog window will present links that bring you to the gene family web pages fig B 5b to close the window click the x at the top right or hit esc B 4 Web interface CoExpNetViz is also available as a web tool on http bioinformatics psb ugent be webtools coexpr index php After running the analysis on the web interface the output files can be downloaded and imported into Cytoscape manually The input form on the web interface is very similar to the plugin but currently as of June 8 2015 the web interface is still a work in progress it is therefore not discussed here This section might be completed in the future B 5 File formats Example bait genes files gene expression files and gene family files can downloaded from the CoExpNetViz website Bait genes For the bait genes file a plain text file is expected with gene ID s separated by any kind of whitespace For information about supported gene identifiers click the ID s button 11 in fig B 1 Gene expression files The gene expression files should have a matrix format with all gene ID s in the first column and all conditions in the first row Columns must be tab separated The expression data must be normalized and summarized this means the data must be background corrected normalized across conditions and summarized to gene
25. her parts of the app When an object has a reference to the CyAppManager it can get information about the application state trough the CyModel and it can invoke Cytoscape actions trough the CyServices Running the coexpression analysis The runAnalysis method in the CyAppManager is responsible for collecting the data specified by the user sending it to server getting back the response and finally displaying the network with the correct layout and visual style Uploading files to the server happens trough a http POST which expects multipart mixed form data The fields of the form data are e baits plain text containing the baits separated by whitespace e matrix0 j matrixl j matrix2 matrix3 and matrix4 the gene expression files e positive_correlation a decimal number in plain text specifying the positive cutoff value e negative_correlation a decimal number in plain text specifying the negative cutoff value e orthologs0 orthologsi orthologs2 orthologs3 and orthologs4 the gene family files APPENDIX A DEVELOPMENT OF COEXPNETVIZ oosusoutruom m 10 Comparative Genomics Co expression Networks The connection with the server is handled by the ServerConn class in the package be samey io The multipart entity is sent to http bioinformatics psb ugent be webtools coexpr index php with two additional url parameters __controller api and __action execute_job This al lows the server side program to differentiate betw
26. icated apple Malus x domestica Borkh Nature genetics 42 10 833 9 October 2010 ISSN 1546 1718 doi 10 1038 ng 654 URL http dx doi org 10 1038 ng 654 Zhang W Liu T Ren G Hortensteiner S Zhou Y et al Chlorophyll degra dation the tocopherol biosynthesis related phytol hydrolase in Arabidopsis seeds is still missing Plant physiology 166 1 70 9 September 2014 ISSN 1532 2548 doi 10 1104 pp 114 243709 URL http www pubmedcentral nih gov articlerender fcgi artid 4149732k amp tool pmcentrezkrendertype abstract BIBLIOGRAPHY
27. ig B 1 or alternatively upload a file with bait genes 3 in fig B 1 see also appendix B 5 for file formats The next step is choosing gene expression datasets to enter a dataset click the browse button 4b in fig B 1 and navigate to the file for the Species field 4a in fig B 1 any name can be chosen To add additional datasets click the Add species button 5 in fig B 1 a maximum of five species can be used at once Then choose PCC cutoff values 6 in fig B 1 to include only positive or negative correlations set the negative cutoff to 1 0 or positive the cutoff 1 0 respectively If you which to save the output of the analysis check the box Save output 7 in fig B 1 the output will be saved as a tar gz archive in the specified folder The name of the file will be the Title 1 in fig B 1 Finally to run the analysis click the Run analysis button 8 in fig B 1 CoExpNet Viz will then run the analysis on the web server Depending on your internet speed and the size of the gene expression files this can take about ten seconds up to a few minutes When the analysis is complete a new network will appear in Cytoscape Specifying custom gene families In addition to the options described above custom gene families can be used to define orthologous genes By default CoExpNetViz will use PLAZA families to find orthologous genes but by specifying your own orthologous genes other gene families such as Ortho MCL families
28. ion and function evolution through comparison of co expression networks of plants Frontiers in Plant Science 5 August 1 9 2014 ISSN 1664 462X doi 10 3389 fpls 2014 00394 URL http www frontiersin org Plant_Science 10 3389 fpls 2014 00394 abstract Kanehisa M Goto S Sato Y Kawashima M Furumichi M et al Data informa tion knowledge and principle back to metabolism in KEGG Nucleic acids research 42 Database issue D199 205 January 2014 ISSN 1362 4962 doi 10 1093 nar gkt1076 URL http www pubmedcentral nih gov articlerender fcgi artid 3965122 amp tool pmcentrez amp rendertype abstract Li L Stoeckert C J and Roos D S OrthoMCL identification of ortholog groups for eukaryotic genomes Genome research 13 9 2178 89 September 2003 ISSN 1088 9051 doi 10 1101 gr 1224503 URL http www pubmedcentral nih gov articlerender fcgi artid 403725 amp tool pmcentrezkrendertype abstract Maere S Heymans K and Kuiper M BiNGO a Cytoscape plugin to assess overrep resentation of gene ontology categories in biological networks Bioinformatics Oxford England 21 16 3448 9 August 2005 ISSN 1367 4803 doi 10 1093 bioinformatics bti551 URL http www ncbi nlm nih gov pubmed 15972284 Mewalal R Mizrachi E Mansfield S D and Myburg A A Cell Wall Related Proteins of Unknown Function Missing Links in Plant Cell Wall Development Plant and Cell Physiology 55 6 1031 1043 March 2014 ISSN 0032 0781
29. j UNIVERSITEIT GENT Supplementary file 1 Comparative Genomics Co expression Networks Authors Sam De Meyer amp Dr Oren Tzfadia June 8 2015 Contents Contents i List of Figures iii List of Tables v Glossary vii 1 Introduction 1 1 1 Function prediction via coexpression analysis 0 1 1 2 Overview of used computational techniques 04 3 A Development of CoExpNet Viz 5 A 1 Compiling CoExpNetViz D ori ee 2 Ride Pea ome DPA UAE ag 6 A 1 1 Setting up a development environment 6 A 1 2 Compiling the source code 0 ven ede ee et eee Bee 6 A 1 3 Running and testing CoExpNetViz 7 A12 General outline ergy ass nee ay as edhe eek a Eades a a Re ke A A peed Pee ge eG 8 A 2 1 Basic structure of the application 0 8 A 2 2 The GUI classes o oo Be Een He Be ee DE eh 10 A 3 Guidelines concerning Cytoscape plugin development 10 A 3 1 Including other java packages o 000004 10 A 3 2 Adding third party libraries 11 A 3 3 Adding JUnit tests ds az e ls e a e 12 A 3 4 Adding a layout algorithm 2 3 24 at daa nd 15 A 4 Development workflow ars er wre a ee 19 A 4 1 Debugging the plugin code ern ack aa Wera a A EM 20 A 4 2 Stepping into the Cytoscape core program source code 21 B CoExpNetViz user manual 23 Bal Installation sei ienet gana er a a 24 B 2 Usage of CoExpNetViz Karte ae er ee
30. loring microarray data https genevestigator com gv e PLAZA Used for defining gene families in the CoExpNet Viz Cytoscape plugin http bioinformatics psb ugent be plaza CHAPTER 1 INTRODUCTION Appendix A Development of CoExpNetViz Foreword This appendix contains guidelines to continue the development of the CoExpNetViz Cy toscape interface There are two reasons why we included this appendix Firstly as we strongly believe that open source code is great way to share and make scientific discoveries accessible to as many researchers around the world we decided to include a detailed manual for developers This manual will also help us in future extension efforts of developing the CoExpNetViz Cytoscape plugin Secondly the Cytoscape wiki is far from perfect Many code snippets on the wiki are incomplete many are out of date and thus plain wrong and most of the Cytoscape Application Programming Interface API is not documented This chapter contains some general guidelines and howto s that took our developers a lot time to find out We have the intention to post some of these guidelines on the Cytoscape wiki to prevent other programmers from running into the same problems as we did 6 Comparative Genomics Co expression Networks A 1 Compiling CoExpNetViz A 1 1 Setting up a development environment CoExpNetViz just as Cytoscape is being developed at Github see https github com The source code of CoExpNetVi
31. ly install and load plugins as well as other parts of Cytoscape The console gives access to many other aspects of Cytoscape in addition java exceptions are printed to the console Cytoscape can be run from the command line as follows change x to the appropriate version make sure cytoscape sh is executeable Cytoscape_v3 x x cytoscape sh It is important to know that Cytoscape is a collection of smaller parts called bun dles which can be loaded and replaced while the programming is running This is possible trough the Open Services Gateway Initiative OSGi framework Cytoscape is essentially nothing more than a collection of OSGi bundles and every plugin by itself is also a OSGi bundle Incorrect configurations for the apache felix plugin in the pom xml APPENDIX A DEVELOPMENT OF COEXPNETVIZ 8 Comparative Genomics Co expression Networks appendix A 3 1 can result in a failure to activate the compiled plugin in Cytoscape If this is the case run the command list in the Cytoscape console this will show all bundles their ID and their state Then run the command start x replace x with the ID of the plugin to manually start the plugin this will print an error message giving information about why the plugin was not activated More detailed information is also written to a log file located at replace CYTOSCAPE_HOME by the cytoscape intallation folder CYTOSCAPE_HOME CytoscapeConfiguration 3 framework cytoscape lo
32. n eed 24 B 3 The family wise coexpression graph o o 25 B4 Webinterface 1 8 2 nee ee a a 25 Bios Eile formats ds 5 ee AI AL A ible a ARA ae 25 Bb How Lo submit DUES non rear ehe 29 Bibliography 31 List of Figures Al B 1 B 2 B 3 B 4 B 5 Attaching the NetBeans debugger to CytoScape 21 CoExpNetViz input form ef re Be A A RAE 27 Overview of a family wise coexpression graph DE Detail of the gene families ds era wer Eee Bela wlan a 28 Detail of the bait genes o nn nn 28 Linking gene families to the PLAZA website 28 ill List of Tables 1 1 Comparison of community accessible coexpression based gene discovery tools 2 Glossary API Application Programming Interface vii 17 21 22 31 bait gene A gene of interest in the family wise coexpression graph for which you would like to discover conserved coexpression links vii viii 3 7 9 10 BAR the Botany Array Resource 10 CSB DB the Comprehensive Systems Biology DataBase 10 Cytoscape An open source software project for viewing and analyzing high throughput biological data Cytoscape provides functionality for transforming high throughput biological data into networks graphs Network statistics and analysis of these graphs can be done from within Cytoscape In addition Cytoscape provides an Application Programming Interface API for writing plugins to extend the core program s fu
33. nation see http felix apache org documentation subprojects apache felix maven bundle plugin bnd html A 3 1 Including other java packages When a new plugin is created from scratch using the org cytoscape archetypes cyaction app Maven archetype see Creating_an_OSGi_Bundle_Cytoscape_3_App in the Cytoscape wiki a basic pom xml file is created In the plugins tag of this file the OSGi setup is specified by the apache felix plugin The default configuration looks like this lt plugin gt lt groupId gt org apache felix lt groupId gt lt artifactId gt maven bundle plugin lt artifactId gt lt version gt 2 3 7 lt version gt lt extensions gt true lt extensions gt lt configuration gt lt instructions gt lt Bundle SymbolicName gt bundle symbolicName lt Bundle SymbolicName gt A 3 GUIDELINES CONCERNING CYTOSCAPE PLUGIN DEVELOPMENT 10 11 12 13 14 11 N outkrkum m Plant Systems Biology department VIB UGent 11 lt Bundle Version gt project version lt Bundle Version gt lt Export Package gt bundle namespace lt Export Package gt lt Private Package gt bundle namespace internal lt Private Package gt lt Bundle Activator gt bundle namespace internal CyActivator lt Bundle Activator gt lt instructions gt lt configuration gt lt plugin gt Two tags in this plugin are especially important the Export Package and the Private Package tags The first tag specifies which packages
34. nctionality Shannon et al 2003 Saito et al 2012 3 5 7 9 10 17 0994 95 5731 family node A node in the family wise coexpression graph that represents a set of genes which belong to the same gene family Every gene in a family node must be coex pressed with at least one bait gene vii 7 9 family wise coexpression graph A graph where nodes are either homologous gene families or bait genes An edge between a bait gene and a family node represents a coexpression relationship between the bait gene and at least one other gene present in the family node vii 9 GEO Gene Expression Omnibus 3 GO Gene Ontology 2 Highest Reciprocal Rank A coexpression similartiy metric used by the PlaNet database Mutwil et al 2011 based on the reciprocal rank of the Pearson Correlation Coef ficient PCC between two genes See http aranet mpimp golm mpg de faq html for more information vii 4 HRR Highest Reciprocal Rank 4 Glossary Highest Reciprocal Rank IDE Integrated Development Environment 29 30 JDK Java Development Kit 18 vu OSGi Open Services Gateway Initiative 19 21 29 PCC Pearson Correlation Coefficient vii 4 7 10 11 PLAZA A resource for plant comparative genomics Trough a web interface compar ative genomic data of 37 plant species is provided this includes Structural and functional annotations classification into gene families phylogenetic trees and in formation about genome organization among other things
35. networkView getNodeView node setVisualProperty APPENDIX A DEVELOPMENT OF COEXPNETVIZ 45 46 47 48 49 HovonsoutkuoumHm E 13 14 oo ousoautkuwrnH 18 Comparative Genomics Co expression Networks BasicVisualLexicon NODE_Y_LOCATION y where x and y are double primitives The third and last class is SomeLayoutContext this class contains the options for the layout task By adding the Tunable annotation the options will appear in the Cytoscape layout settings window where the end user can change their values A minimal example package com something package_name import org cytoscape work Tunable public class SomeLayoutContext the values specified here are the default values Tunable description A String option public String someStringOption string_option_value Tunable description A double option e g to set minimal distance between nodes public double nodeMinSpacing 20 0 ad as many options as you like Whenever an end user uses a layout algorithm an instance of this context class with the values as specified by the end user is passed to the constructor of the layout task This context object can then be used in the layout task code to retrieve the options Finally to add the layout to the Cytoscape layouts menu put the following code in the CyActivator start method Also make sure that the package containing the layout code is included in the Export Package
36. ntegrate it in the plugin which is not user friendly Maybe the best solution would be to reprogram the C core in java We opted for the second best solution to run the C core on a web server An advantage of this approach is that a user can also make use of the web server by itself to run the analysis without the requirement of any installed programs lowering the barrier to try out the tool Another advantage is that by using a web server big gene family files do not have to be included in the plugin jar lowering the download size The standalone web application is available at http bioinformatics psb ugent be webtools coexpr A 2 1 Basic structure of the application The Cytoscape plugin code is divided into a number of packages These packages to gether with embedded dependencies form an OSGi bundle which can be loaded into Cytoscape The starting point of the application is CyActivator java which extends AbstractCyActivator in the package be samey internal When Cytoscape is loaded the start method is invoked any services the plugin provides are specified in this method There are three services provided by CoExpNetViz the app GUI which is located in the A 2 GENERAL OUTLINE Plant Systems Biology department VIB UGent 9 Apps menu a layout algorithm and an event listener to run code on certain events that happen in the core program The menu action which launches the CoExpNetViz GUI is invoked by the actionPerfomed me
37. ode and step trough the code line by line every variable value can checked and watches can be created to to check how one or more variables changes while the program is running If only a very limited amount of source code was changed clicking the Apply Code Changes button will apply the effect of the change immediately without even recompiling the plugin This can be very useful for processes that require many trial and error runs as for example placing GUI components in aesthetically pleasing positions But whenever a method definition is changed added deleted Apply Code Changes will not longer work and recompiling the app will be necessary After recompiling the Netbeans debugger can point to incorrect line numbers while stepping trough source code This can be solved quickly be stopping the debugging session Cytoscape will just keep running while waiting A 4 DEVELOPMENT WORKFLOW Plant Systems Biology department VIB UGent 21 D t Attach Debugger Java Debugger JPDA Connector SocketAttach Attaches by socket to other VMs Transport dt_socket Host samArch Port 12345 Timeout ms Cancel Figure A 1 Attaching the NetBeans debugger to CytoScape This window can look different depending on which operating system is used for a new debugger to attach Then re attach the debugger in Netbeans and continue working as before A 4 2 Stepping into the Cytoscape core program source code The Cytoscape A
38. omparative Genomics Co expression Networks Accessing resources is usually done with the getResource method of some class loader When using the default maven layout as described above any test resources will au tomatically be added to the search path of the getResource method An example of getting a test resource SOI LES menor im A TEHE Class public void testSomeMethod only works if the default Maven file structure is used e g the the resource used here is located at PLUGIN_ROOT src test resources someResource foo URL url getClass getClassLoader getResource someResource foo InputStream is url openStream now the contents of the resource can be accesed with is Using Cytoscape objects in JUnit tests Add the following to the pom xml to use Cytoscape objects in JUnit tests lt dependency gt lt groupId gt org cytoscape lt groupId gt lt artifactId gt model impl lt artifactId gt lt version gt 3 1 0 lt version gt lt type gt test jar lt type gt lt scope gt test lt scope gt lt dependency gt lt dependency gt lt groupld gt org cytoscape lt groupld gt lt artifactId gt model impl lt artifactId gt lt version gt 3 1 0 lt version gt lt scope gt test lt scope gt lt dependency gt Then to get an instance of a CyNetwork object for example package com something package_name import org cytoscape model CyNetwork import org cytoscape model NetworkTe
39. r could apply GO enrichment using BiNGO for example Maere et al 2005 or analyze network properties utilizing other Cytoscape plugins By providing a user friendly Cytoscape plugin it our goal to make gene function predic tion trough comparative transcriptomics analysis accessible to plant researchers without specialized bioinformatics knowledge or programming skills 1 1 FUNCTION PREDICTION VIA COEXPRESSION ANALYSIS Plant Systems Biology department VIB UGent 1 2 Overview of used computational techniques Programming techniques e Java OpenJDK Writing the CoExpNetViz Cytoscape plugin http openjdk java net Git GitHub Collaboration and version control http git scm com and https github com Maven Building documenting and organization of the plugin http maven apache org OSGi Integrating the CoExpNetViz tool into the Cytoscape core program http www osgi org e Perl BioPerl amp Python Several small scripts to parse files into the desired format https www perl org http www bioperl org and https www python org eR The first version of CoExpNetViz was written in R http cran r project org Bioconductor Downloading manipulating gene expression datasets http www bioconductor org geplot2 Creating graphs for this document http ggplot2 org Tools used e Cytoscape Analyzing coexpression data http www cytoscape org e Genevesigator Exp
40. rack or revert changes NetBeans also comes with a graphical debugger and built in mechanism for executing unit tests among many other features Importing a Maven project into Netbeans To import a Maven project click File Open Project and navigate to the folder containing the maven project Select the folder and click Open Project NetBeans will automatically detect that the project is a maven project and if a git directory is present NetBeans will automatically integrate with Git Code compile test Start Cytoscape from the command line also open a tail for the Cytoscape log file in another terminal window This enables control over which OSGi bundles are activated and gives information about eventual exceptions errors and OSGi APPENDIX A DEVELOPMENT OF COEXPNETVIZ N 20 Comparative Genomics Co expression Networks output Print statements to STDOUT from bundles are printed to the console whenever a new bundle is activated it is printed to the log file Cytoscape_vx x 0 cytoscape sh in another terminal window tail f CytoscapeConfiguration 3 framework cytoscape log Also as mentioned in appendix A 1 3 symlinking the jar file to the Cytoscape apps folder will make Cytoscape to automatically update the plugin whenever it is recompiled Thus it is almost never needed to restart Cytoscape while working on a plugin In s PLUGIN_ROOT CoExpNetViz target CoExpNetViz 1 0 SNAPSHOT jar CYTOSCAPE_HOME CytoscapeConfigura
41. should be exported by the bundle any packages that are not exported are unavailable at runtime To export all the packages of the plugin during runtime change the line to lt Export Package gt bundle namespace lt Export Package gt The second tag Private Package specifies which packages should not be exported dur ing runtime but should still be included in the bundle The Export Package tag takes precedence over Private Package tag A 3 2 Adding third party libraries To add a dependency to the plugin add to following lines to the pom xm1 in the instructions tag of the felix plugin configuration lt Embed Dependency gt scope provided lt Embed Dependency gt lt Embed Transitive gt true lt Embed Transitive gt lt Import Package gt resolution optional lt Import Package gt The Embed Dependency tag specifies which jars should be embedded in the plugin By using the character all jars in the dependencies tag of the pom xml are included this means the Cytoscape APIs will also be included which is not necessary To prevent this scope provided is added now all jars will be included except for the ones which have the provided scope All Cytoscape API jars are provided by the Cytoscape program therefore all Cytoscape API jars must be marked as provided in the pom xml by adding lt scope gt provided lt scope gt as follows lt dependency gt lt groupId gt org cytoscape lt groupId gt lt artifactld gt s
42. stSupport JUnit imports and other packages if needed public class ClassNameTest Constructor setUp tearDown etc x Some west method in a test class Er public void testSomeMethod NetworkTestSupport nts new NetworkTestSupport CyNetwork cn nts getNetwork now you have cn a CyNetwork instance to do some tests with A 3 GUIDELINES CONCERNING CYTOSCAPE PLUGIN DEVELOPMENT 18 19 DAUNE NOOR WNEH ONaKRWNEH hip Ne oo ANow1 rwner Plant Systems Biology department VIB UGent 15 At some point while running tests a NoClassDefFoundError might occur for example org cytoscape event DummyCyEventHelper java lang NoClassDefFoundError at at at This means that some class that was available at compile time is not longer available while running tests In this example the missing class is DummyCyEventHelper To resolve this problem add the package containing the missing class as a test depency lt dependency gt lt groupId gt org cytoscape lt groupId gt lt artifactId gt event api lt artifactId gt lt version gt 3 1 0 lt version gt lt type gt test jar lt type gt lt scope gt test lt scope gt lt dependency gt A 3 4 Adding a layout algorithm First of all two Cytoscape dependencies are required lt dependency gt lt groupId gt org cytoscape lt groupId gt lt artifactId gt layout api lt artifactId gt lt version gt 3 1 0 lt version gt lt scope gt pro
43. tag of the Apache felix setup in the pom xml see appendix A 3 1 UndoSupport undoSupport getService context UndoSupport class SomeLayout layout new SomeLayout undoSupport Properties layoutProperties new Properties layoutProperties setProperty preferredTaskManager menu TITLE is inhereted from ServiceProperties by AbstractCyActivator and is a String with value title The toString method is defined in AbstractLayoutAlgorithm in the Cytoscape layout api and returns the human name of the layout algorithm see above Thus the line below sets the layout name in the menu to the human name layoutProperties setProperty TITLE layout toString layoutProperties setProperty PREFERRED_MENU name of menu entry MENU_GRAVITY is inhereted from ServiceProperties as well The higher the number the lower the entry appears in the menu layoutProperties setProperty MENU_GRAVITY 10 registerService bundleContext layout CyLayoutAlgorithm class layoutProperties A 3 GUIDELINES CONCERNING CYTOSCAPE PLUGIN DEVELOPMENT DAUNE 10 11 12 13 Plant Systems Biology department VIB U Gent 19 To apply the layout programatically get the references to following objects in the start method of the CyActivator and pass them to the class that should apply the layout algorithm TaskManager taskManager getService context TaskManager class C
44. thod inherited from AbstractCyAction of the MenuAction class To summarize there are two important entry points into the code firstly there is the CyActivator class whose start method is run when Cytoscape is started Secondly there is the MenuAction class whose actionPerfomed method is run when the user clicks the CoExpNetViz entry in the Apps menu The CyAppManager CyModel and CyServices There are three other classes of interest in the be samey internal package e One instance of CyAppManager is created in the CyActivator start method and is central to the plugin This class acts a as central control point it provides methods to get the settings directory run the coexpression analysis on the web server and to get IO helper classes In addition the CyAppManager has references to the CyModel and the CyServices The same CyAppManager instance is passed on to many parts of the plugin e The CyModel is passed as an argument to the constructor of the CyAppManager The CyModel has fields with getters and setters which are used to keep track of the application state e The CyServices class is also passed as an argument to the CyAppManager constructor This class has fields containing the Cytoscape model classes along with getters and setters for those fields Instances of the Cytoscape model classes can only be obtained by the CyActivator by wrapping all these model classes in a CyServices object they are more easily passed around to ot
45. tion 3 apps installed When clicking the Clean And Build button or pressing shift F11 the Maven goal clean install is executed and Maven output from executing goals is printed to an output window in the IDE Thus in a typical workflow you modified some code then hit shift F11 and see if the result is as expected in Cytoscape this way you can check the effect of the code changes in a matter of seconds To test the currently open source file hit shift F6 or to execute all JUnit tests hit alt F6 If there are test files present in the default testing directory see appendix A 3 3 the maven surefire plugin will perform the JUnit tests When using version 2 15 or higher of the surefire plugin individual methods can be tested as well This provides a quick way to test out code without having to use Cytoscape A 4 1 Debugging the plugin code Cytoscape can started from the command line with the option debug this will print the following output to the terminal Cytoscape_vx x 0 cytoscape sh debug Listening for transport dt_socket at address 12345 Then in NetBeans click Degug Attach Debugger this will open a dialog asking for a socket address Choose the options as shown in fig A 1 except for the hostname which is the name of the computer After clicking OK Cytoscape will start and every thread started by Cytoscape can be seen in the IDE window In addition it is possible to place breakpoints in the plugin source c
46. to gain information about the coexpression of enzymes and regulators correlated with the presence or absence of that metabolite Usadel et al 2009a Rhee and Mutwil 2014 This way genes of unknown function that are co regulated with proteins that are known to be involved in the production of the metabolite of interest can be proposed as new candidates for experimental validation Comparative transcriptomics The practice of combining and integrating expression data from multiple species known as comparative transcriptomics adds another layer of information to transcriptome analysis increasing the predictive power There are a few problems associated with transcriptome data one of them being that it is inherently noisy increasing the number of both false positives and false negatives Another problem is the failure to cover all genes increasing the number of false negative results as well The assumption used in comparative genomics is that by looking at expression patterns that are conserved between orthologous genes the noise is reduced while true regulatory interactions gain strength as they are conserved across species And indeed several studies have shown that this is the case Bergmann et al 2004 Stuart et al 2003 Visualization and interpretation Comparative coexpression analysis often results in long lists genes that have varying degrees of similarity to each other in terms of expres sion patterns or sequence A network grap
47. vided lt scope gt lt dependency gt lt dependency gt lt groupId gt org cytoscape lt groupId gt lt artifactId gt work api lt artifactId gt lt version gt 3 1 0 lt version gt lt scope gt provided lt scope gt lt dependency gt Then a layout class must be created which extends AbstractLayoutAlgorithm the class should contain at least the methods shown below package com something package_name import java util HashSet import java util Set import org cytoscape model CyNode import org cytoscape view layout AbstractLayoutAlgorithn import org cytoscape view model CyNetworkView APPENDIX A DEVELOPMENT OF COEXPNETVIZ 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 16 Comparative Genomics Co expression Networks import org cytoscape view model View import org cytoscape work TaskIterator import org cytoscape work undo UndoSupport public class SomeLayout extends AbstractLayoutAlgorithm Creates a new SomeLayout object param undo Ys public FamLayout UndoSupport undo 4 the two Strings here are the name that can be used to refer to the layout algorithm from source code and the name that will appear in the layouts menu in Cytoscape respectively super computer_layout_name human_layout_name undo
48. yLayoutAlgorithmManager cyLayoutAlgorithmManager getService context CyLayoutAlgorithmManager class Use the following code snippet to eventually apply the layout get a reference to the network you which to lay out CyNetworkView cyNetworkView gt the node attribute used to apply the layout with String attrName SomeLayout layout SomeLayout cyLayoutAlgorithmManager getLayout computer_layout_name TaskIterator ti layout createTaskIterator cyNetworkView below the default options are used but you can change some values of the context first layout createLayoutContext CyLayoutAlgorithm ALL_NODE_VIEWS attrName taskManager execute ti A 4 Development workflow This section contains some suggestions on how to optimize the code compile test cycle for writing Cytoscape apps These suggestion are not the best way to do it but they work well for us We have used the NetBeans Integrated Development Environment IDE https netbeans org which is a popular IDE for java Another maybe more popular choice is the Eclipse IDE https eclipse org Both IDEs are very similar so suggestions here are probably equally valid for both NetBeans and Eclipse Using NetBeans has many advantages since NetBeans is aware of the Maven directory structure and dependency mechanism NetBeans also integrates with Git and marks changes since the last git commit in the code line numbers making it really easy to t
49. z can be found at https github com CoExpNetViz git the Cytoscape source code is also available on Github but is not required for building plugins although it can be helpful see appendix A 4 2 Several programs have to be installed to develop a Cytoscape plugin First of all the Java Development Kit JDK has to be installed Cytoscape runs on java 1 6 and java 1 7 Java 1 8 might appear to work as well on first sight but running Cytoscape on java 1 8 can cause many hard to track bugs that only appear after using it for a while It is therefore recommended to use JDK 1 6 or JDK 1 7 The second required program is Maven This is a framework that helps in organizing and compiling a java project and Cytoscape itself is also built using Maven It can be downloaded from http maven apache org At least Maven 3 0 is required to success fully compile the code The third required program as mentioned before is Git It is recommended to use the most recent version Git can be downloaded from http git scm com The fourth and most obvious program is Cytoscape itself although it should be noted that it is technically not required to have Cytoscape installed to write and com pile a plugin CoExpNetViz is written for version 3 1 therefore to test the plu gin this version or higher should be installed The program can be downloaded from http www cytoscape org A 1 2 Compiling the source code To get a local repository copy of the source
Download Pdf Manuals
Related Search