Home

CyKEGGParser User Manual - The Institute of Molecular Biology

1. where the set of Entrez IDs for each node is kept After performing protein protein interaction based tuning and drill down the Entrez ID for each gene is kept in entrezld attribute Alternatively the user may load another attribute data where the identifiers are kept The attribute should be selected and set from the Select attribute containing Entrez gene ID drop down menu EntrezlDs attribute is set as default Additionally in Select attribute specifying entity type menu the user should also select the entity type to be processed by gene expression based pathway tuning For this they should choose the attribute name and its value s e g if gene entity type is chosen only the nodes having gene value of the attribute Type will be considered for removal during the tuning process Tissue and threshold specification After specifying the gene expression data source the tissues contained in the dataset are listed in the Select the tissue drop down menu After choosing the tissue the Load dataset button is activated clicking which will result in loading the gene expression data for the specified tissue and the nodes in the network according to chosen attributes After a while the data will be loaded and the user will be able to set gene expression threshold The latter is based on percentiles of the distribution of gene expression values for the specified tissue Along with those the absolu
2. KGML and BioPAX_level2 it will assign maplink type to those interactions where one of the nodes is of type Map and will give the default value PPrel otherwise in case of BioPAX_level3 the interaction will be removed from the network since the latter format does not allow for interactions between non physical entities Be aware that because of concurrency issues the pathways may sometimes fail to convert to BioPAX format in this case try converting it in another format and or using Keggtranslator independently Saving back to KGML format allows for using modified and corrected KGML pathways in other applications and analyses The report of the results is kept in logs parsing log file in the app s directory by default this is SUSER_HOME CytoscapeConfiguration app data CyKEGGParser Future developments Metabolic pathway parsing In contrast to other pathway types KGML files of metabolic pathways contain lt reaction gt tags which contain information about substrate and product compound transitions occurring with participation of an enzyme At the moment CyKEGGParser does not rely on these interaction tags but uses only lt relation gt s In some cases this causes disruption of interactions between pathway nodes These however are fixed during pathway drill down In future releases CyKEGGParser will be adjusted for correct handling of lt reaction gt tags Application of pathway signal flow algorith
3. important to take into account the source of information on interactions depicted in pathways and be able to filter among real physical and indirect interactions as well as experimental and prediction based sources All of these may have crucial effects on various types of automated pathway based analysis algorithms and therefore should be handled appropriately Therefore we have implemented functionality for customizing pathways according to specific biological context particular tissue or cell type and experimentally confirmed physical interactions To perform pathway tuning go to Apps gt KEGGParser gt Pathway tuning From the top drop down menu Select network choose the pathway to be tuned In the bottom of the panel choose the tuning type gene expression or protein protein interaction as well as choose the tuning mode generate new network or modify current network Specify corresponding tuning settings and press the Tune button Select network 52B cell receptor signaling pathway Gene Expression Settings _Protein protein interaction Settings Data source BioGPS only for human genes Select attribute containing Entrez genelD 6 User supplied data EntrezIDs Mo file chosen Select the tissue B lymphoblasts Select attribute specifying entity type Load dataset Set expression threshold percentile abs value 16 Tuning mode Generate new network Protein
4. v Pal V MINT Select attribute specifying entity type et ppi threshold J 0 100 200 300 400 500 600 700 200 900 1000 Tune the network based on Tuning mode Gene expression Generate new network Change current network Save settings Cancel The PPI tuning performs by initially drilling down the pathway through expanding each node of gene type into separate nodes for each gene The gene node label equals to Entrez ID of the gene Furthermore the algorithm iterates on all the pairs of interacting nodes and connects those member gene nodes for which there is real physical interaction The existence of a physical protein protein interaction is based on the data provided by String database http string db org which contains data from GRID DIP KEGG MINT and PDB databases The algorithm queries the database for the interactions between genes in the pathway according to data sources and confidence score provided by the user In String the confidence score is derived by combining evidence about protein protein interactions from various sources adjusted for probability of randomly observing the interaction More information about confidence score meanings and interaction sources can be found at the Help page of String database http string db org newstring cgi show info page pl Userld z7Cu7ePjQxXnl amp sessionld rD f8EsSHGmAQ The interactions are manually updated in the local My SQL database and the version of S
5. will appear in Cytoscape as an empty node the group NFKA node and member nodes separately The edge CHUK NFKBIA distribution for these nodes varies and depends on the KGML file If Process group nodes option is chosen during parsing the group node will disappear and its member nodes will be connected with binding edges and arranged in a sequence The order of node arrangement depends on the incoming and outgoing relations of those nodes This processing is important for preserving pathway A CARD11 BCL10 MALT1 and CHUK IKBKG IKBKB form two groups flows which may have significant effects on which according to the Kam file appear echalatel from their group pathway flow based algorithm results nodes n a B After corrections group nodes disappear and the member nodes are connected Protein compound protein PCP interaction processing Interactions of this type are represented as protein protein interactions with compound interaction subtype For example in the KGML file of Chemokine signaling pathway the interaction between nodes PIK3R5 and AKT3 and PIK3R5 and PREX1 is A a i presented as ooo E lt relation entry1 10 entry2 57 type PPrel gt hig lt subtype name compound value 62 gt P lt subtype name activation value amp gt gt lt relation gt lt relation entry1 10 entry2 45 type PPrel gt lt subtype name compound value 62 gt lt s
6. 929 20 3238 20 8381 19 0262 20 7 3007 0 5007 6 5500 2 3007 Tissue specific tuning options If gene identifiers are repeated in the dataset they are handled in the manner specified by the user in Preferences In order to change gene conflict handling settings go to Apps gt KEGGParser gt Preferences and choose one of the three modes mean minimum or maximum If the mean mode is chosen gene expression values will be averaged for the same gene In case of minimum or maximum mode the minimum or maximum values will be chosen among single gene expression values for a particular tissue Those genes whose expression values are not found in the dataset may either remain untouched in their nodes or removed from the network upon user s choice This option can be adjusted by going to Apps gt KEGGParser gt Preferences and checking unchecking the Keep the genes with missing values in the nodes checkbox V Process protein compound protein interactions Correct binding interaction directions Tissue specific tuning options Multiple value handling in expression data Mean gt Minimum Maximum V Keep genes with missing values in the nodes PPI based drill down options Keep indirect interactions in the network Save preferences as default OK Cancel Attribute specification panel After CykKEGGParser parses KEGG pathways it creates a node attribute in Cytoscape named EntrezlDs
7. CyKEGGParser User Manual Table of Contents WO CAC EO ea E E E 3 VSO E e A E E EE ET 3 CRIO a E ENE A AEE E A E A 3 PMC NUS AE EE A T A A E T E T E A A eee ence a E ete eee eee et 3 SS EN E E E e EE S A A E A E E EA A A T E E 4 PEM Waly IO A GINS aiaia A E E EE A AE 4 Laoding KEGG pathways from local KGML files ce cccecceccceceeeeseeesecsscessesseessessssessesseessaeansuesaaesaecsaesaeeeeeeeeeeees 4 Importing KEGG pathways from the WeD uu cccccsecceecceecececeseeeesecesecseeessesseessesesesseneeesngguauesaaesaeesaesseesaeeeeeeees 4 EETA EE I E A E EE EE EE E E E EA A eee ee 4 Automatic correction OptionS ssssssessssssesrsssrrersssssrersesrssrerssssroressseoressssssesesrerroressreeresssereesessaseresreseeresererressseseereo 5 Group Node proces SiMe cate sos cacti cccnacec E iE 5 Protein compound protein PCP interaction processing essessseeseeresssssssrererereressseossrereresenesssesesrereeesseesssene 6 Correction of binding interaction directions sssssesssesessereseeseeereerereseeesresssessessssssssssessseoseooreeorevrrererrerrerereresee 6 Par ME re UONE a EE ETNE OANA EAN EEE EERE 7 PARWA E a E E E E A A 8 Tissu specific pathway TUNING siccisreneenreisiieiinipidsen eia anaana E A REEERE EE EEE EE ERNEK 9 User supplied gene expression data format essseeseeesseessesseessssssesssossessseosreoreevereeereeerereeeeeeeeseeeseesseeseessesse 9 Gene CONTICE NANGIING scseciecevecsaccascsccsccadeadstarvaisss
8. anesaevacvandeadeguaedievavesseseasiasbadss Error Bookmark not defined Attribute specification DANE sce sscsesedense see teacteacsenteigeidevduaeucnestastassacuiasadasabuaisaigetaeidesses veussdessesosbeveendewseteapenantats 10 Tissue ANA threshold SHCCITICATION siscisssssscesvscnssadedcrdeateosesecsecasecansatessnssaavaasnddavdavsscevedsvededuaton dob edcsecsesdesauseatocads 10 T ninge results TOR CAs ar sacs acestesseceivancasacaacesdeatcostenceccteatidattatatcnas bean ataniaatpaiuesduseundapdandandosegavausngsbecebeebounaeedansesenass 10 Protein protein interaction based CUNING cccccccccecceecceceececsecssesseeseesssessesesseessaueauansuecaueaaecseceseceeeeeeeeeeeeeeeeees 11 SVINE THE DAWO Senne ee ee ee re ee eee ee ee eee 13 FPUrure deve lOni eee eee te ene ine nr EEEE ne eee en tees ere eee 13 WIELA DONIC pathway NSIC pie ascii cece tae A ccc aa bates Saueees a ae ey aes tee ee 13 Application of pathway signal flow AlZOrith ccccccessssecssesessssessessssesseesuaanaaecaaeseesaaecaeeeeeeeeeeeeeseeeeeeeeeees es 14 And of course bug fixing and improvements cceecssscsseesseseecssessssssscsscsnseaaesaaesaeauaesaeeeeeeeeeeeeeeeeeeeeeeeeeesy 14 Introduction CyKEGGParser v1 0 is an app for Cytoscape which operates on pathways derived from the Kyoto Encyclopedia of Genes and Genomes KEGG pathway database It provides the user with functionality of parsing and visualization of KEGG pathway maps in Cytoscape Alon
9. g attributes in Cytoscape If corrections are applied to nodes and edges these will be mentioned with the attribute Comments Automatic correction options In most cases KGML files do not fully correspond to the static pathway images Inconsistencies may include absence of event or entity labels reversed directions for some associations absence of some interactions ambiguous definitions of group nodes compounds and their interactions v Process protein compound protein interactions CyKEGGParser is able to automatically correct for some of C Correct binding interaction directions those inconsistencies based on meta information contained in KGML files or on pathway topology In other cases though the Se ee E user will have to perform corrections manually using Mean Cytoscape functionality of adding removing and modifying pots Maximum nodes and edges To adjust automatic correction options of go to Apps gt F Save preferences as default KEGGParser gt Preferences and check uncheck the automatic corrections option boxes ox canai Those are explained in detail below Group node processing In KEGG pathways nodes are combined into groups if protein complex formation is a necessary step for downstream events to take place In KGML files there are separate nodes for groups and for each MAPK signaling pathway of the group members and after parsing with Process group nodes box unchecked they
10. g with this it provides an option for semi automatic correction of inconsistencies between KEGG static pathway images and accompanying KGML files The two main features of the app are the functionality of deriving tissue specific pathways and for protein protein interaction PPI based drill down of the pathways Development The app has been developed by the members of the Bioinformatics Group at the Institute of Molecular Biology of the National Academy of Sciences of the Republic of Armenia IMB NAS RA You can visit the group s webpage at the following link Citation When using the app in your research please refer to the app webpage License Copyright C 2013 Lilit Nersisyan amp amp Arsen Arakelyan BIG IMB NAS RA This program is free software you can redistribute it and or modify it under the terms of the GNU General Public License version 3 The license can be found at http www gnu org licenses gpl html Getting started Installing the app from web In Cytoscape go to Apps gt App Manager choose CyKEGGParser under Network generation Network analysis pathway database categories and click on the install button In case of successful installation the plugin menu KEGGParser should appear under Apps menu Pathway loading Laoding KEGG pathways from local KGML files Go to Apps gt KEGGParser gt Load KGML gt Load local KGML and choose the KGML file to be imported Importing KEGG path
11. m Our group has developed pathway signal flow PSF algorithm Arakelyan A Aslanyan L amp Boyajyan A 2013 High throughput Gene Expression Analysis Concepts and Applications Sequence and Genome Analysis II Bacteria Viruses and Metabolic Pathways ISBN 978 1 480254 14 5 iConcept Press Integration of PSF algorithm in CyKEGGParser will allow for calculation of signal flows in KEGG pathways And of course bug fixing and improvements Keep positive there are not many of them
12. protein interaction Change current network Save settings Tissue specific pathway tuning In the tuning window navigate to Gene Expression Settings tab and choose the source of gene expression data For human pathways the user can choose to tune based on the dataset provided by GeneCards http www genecards org which contains BioGPS data representing log transformed and normalized data for human normal tissue and cancer cell lines http biogps org goto welcome Alternatively the user can supply gene expression data in the specified format User supplied gene expression data format The data should be presented in tab delimited format where the first column represents gene identifiers Entrez Gene IDs and the remaining columns are for gene expression values for different tissues or cell types one column per each tissue The first line of the first column should be named ID The first line of each tissue column should contain the tissue header and the rest of the rows should represent gene expression values Missing values should have value 0 0 ID B lymphoblasts Adipocyte AdrenalCortex Adrenalgland 1 4 0000 4 1000 4 1500 4 1000 2 6 1100 120 5200 165 3 700 0 0000 9 of 7222 23 4444 19 5611 32 0944 10 3 1500 4 2500 4 5667 3 0167 12 4 6500 36 6000 7 7000 13 8500 13 T 2333 2 8333 23 3000 21 2000 14 132 7500 20 9000 19 2500 16 0500 16 709 7833 217 60833 169 8833 167 7333 18 12 0000 15 3333 22 0107 14 7500 19 15 6
13. relation entry1 61 entry2 63 type PPrel gt lt subtype name binding association value gt lt relation gt lt relation entry1 61 entry2 62 type PPrel gt lt subtype name binding association value gt lt relation gt These directions are in fact opposite to the real picture of signal flow Automatic corrections of binding interaction directions allow for reversing these edges In some cases however the incorrect edges might still remain in the network thus the user should be careful when directionality matters LS SUS s pathway p EE K10784 CD247 K10785 CD247 Results of corrections of protein compound protein interaction directions Red arrows indicate directions in KGML files green arrows indicate reversed directions Parsing result logging The report of the results is kept in logs parsing log file in the plugin s directory by default this is SUSER_HOME CytoscapeConfiguration app data CyKEGGParser Pathway tuning KEGG pathways are generalized pathways with nodes representing collections of gene products with abstract interactions between them not always occurring together in the same biological context Specifically one of the abstractions is that paralogous genes gene products are grouped into one node Another abstraction is that pathways are drawn under the assumption of a generalized cell in which all the genes are assumed to be expressed Additionally it s
14. te value corresponding to each percentile is displayed in the abs value box Tuning results logging All the genes with expression values higher or equal to the specified threshold will remain in the network after tuning Since each node contains a number of genes the gene ids contained in EntrezlDs attribute will be removed from the network If no gene ID remains in a node the node will entirely be removed from the network If no other entity type is specified aside from gene the other entities such as compound and labels will remain as they are The report of the results is kept in logs tuning log file in the app s directory by default this is SUSER_HOME CytoscapeConfiguration app data CyKEGGParser Protein protein interaction based tuning Go to Apps gt KEGGParser gt Pathway tuning From the Select network drop down menu choose the pathway to be tuned In the tuning window navigate to Protein protein interaction Settings tab Here the user can choose the source of information about interactions and set the confidence score threshold In the bottom menu choose Protein protein interaction tuning type and press the Tune button Select network 52_B cell receptor signaling pathway Gene Expression Settings Protein protein interaction Settings Interaction source STRING 9 05 z S Select attribute containing Entrez genelD Source EntrezIDs x 7 GRID V DIP V KEGG PDB
15. ther remain blank assigned default values or Cytoscape inherent properties values such as color and coordinates in the resulting KGML file see below In addition CykKEGGParser uses KEGGTranslator in order to convert the pathways in BioPAX_level2 and BioPAX_level3 formats http www biopax org documentation php During conversion the network is initially saved as a KGML file which is then given as input to KEGGtranslator The latter is called through command line call with the following arguments java jar KEGGtranslator jar input in_file xml output out_file format out_format KGML format saving assures that all the attributes required for BioPAX translation are available For nodes these are entry id and entry type attributes these are assigned the default values the next maximum id in the network and gene respectively Node color and coordinates are not required for format conversion however if those are missing in the attributes they will be assigned the values they have in Cytoscape For interactions the required attribute is the type attribute If this attribute is missing in the network it is assigned a value based on source and target node types as follows if both nodes are of type gene the interaction is assigned the type PPrel if either of the nodes is of type compound the type PCrel is assigned otherwise the algorithm looks at the required format in case of
16. tring used is mentioned on the Tuning dialogue If the Keep indirect interactions in the network checkbox is checked the indirect interactions will remain in the network if those are not found in the interaction database and will be removed otherwise Note that in some pathways the nodes are duplicated in order to keep the pathway view cozy Drill down combines these nodes into one redirecting corresponding interactions If one of the nodes is not of type gene compounds and entity labels the interactions between expanded nodes will remain as they are Finally all the nodes that appear unconnected will be removed from the network Note that in the drilled down network gene nodes originating from one node in the original network will appear on top of each other The user will have to allocate the nodes apart or apply Cytoscape layouts to see all the genes and their interactions in the network The report of the results is kept in logs tuning log file in the app s directory by default this is SUSER_HOME CytoscapeConfiguration app data CyKEGGParser Saving the pathway In order to save the pathways in KGML format go to Apps gt KEGGParser gt Save network gt Save as KGML All the modifications done to the network including tuning based changes and added or removed nodes and edges are saved in the attributes specific to KGML format If the necessary attributes values are missing in Cytoscape those ei
17. ubtype name activation value amp gt gt lt relation gt where the compound 62 is PIP3 While in the original KGML PIP3 is not connected to any node PCP fixing allows restoring the right order of connections in the final network A Direct protein protein interactions between PIK3R5 and AKT3 and PIK3R5 and PREX1 The compound node is left out from the network B Note that PCP interaction processing is of interest for Protein compound protein interactions between those nodes pathway flow recovery and for better visualization If recovered citer automatic processing instead direct protein protein interactions are of particular concern e g for further network drill down based on protein protein interaction data processing of PCP interactions should be turned off This specifically matters in case of metabolic networks Correction of binding interaction directions Although a binding event is non directional in terms of molecular interactions the direction of an event makes sense in the context of information flows in a graph reverse directions may lead to flow disruptions Direction fix is based on relative positions of two interconnected nodes The node upstream in the pathway is considered as the source while the other is considered as the target In NOD like receptor signaling pathway the binding interactions PDCD1 to PTPN6 and CTLA4 to PTPN6 are represented in the KGML file as follows lt
18. ways from the web Go to Apps gt KEGGParser gt Load KGML gt load KGML from web Note that the window for web import will take a while to load organism and pathway lists From the drop down menu choose the organism and pathway names or alternatively type in the organism identifier or KEGG pathway ID in the left hand boxes Press the Enter key after typing in order for the selection to take place Press the Load button after choosing a pathway to load Pathway parsing Type in KEGG identifier or choose from list After specifying the pathway to l load it is parsed with current Organism hsal Homo sapiens human automatic correction options Pathwa 04662 B cell receptor signaling pathwa z Se applied and visualized in Home sapiens human B cell receptor signaling pathway Load Close Cytoscape environment Note that a visual style named kegg vs is created If the pathway is visualized with another style simply go to Style panel and choose kegg_vs visual style Because of concurrency issues the visual style may fail to be set or may be set incorrectly In this case the user will have to switch or re switch to kegg_vs visual style manually and if necessary uncheck the Lock node width and height option to achieve KEGG specific visual style Along with pathway parsing all the meta data contained in KGML files is mapped onto nodes and edges through creation of correspondin

CyKEGGParser User Manual - The Institute of Molecular Biology

Contents

Download Pdf Manuals

Related Search

Related Contents