Home

Cell Illustrator User Guide

1. Cell Illustrator 1 5 Cell Illustrator 1 5 NES CENAE CERCA PEE ee CENNAIAELRRCON DEE New Canvas 3 New Canvas 3 x E f a x E b gt Pl a C 7 pa x 5 x g 3 E 2 2 S h is S a i 7 ol x fe 01 ial i i a h1 h1 h1 h1 b m SS 6 Q F a 05 o6 o7 08 e e h1 h1 h1 h1 L L 9 m2 010 m2 11 m2 12 m2 9 m1 010 m1 11 m1 12 m1 o9 010 011 012 bf Figure 87 Usage of the hierarchization 2002 2008 GNI Ltd and Human Genome Center Institute of Page g3 of 100 Medical Science The University of Tokyo All rights reserved 8 3 Edit Parts The Edit Parts button is the leftmost button in the Edit toolbar With this button you can register a model file as a reusable item Once registered the model is accessible in the drop down list and the elements in the file can be imported onto a canvas in the same way the command Import Model works With Edit Parts you can organize a repository of frequently reused models Cell Illustrator comes with several pre registered parts to demonstrate this functionality 2 2 20 874 EP 2 oe so View Dialog Window Help i R OOO ml Bes 1 FNRS ER He modell gon e3 __ m3 e3 __ m3 Aos leE em AAE eS ee lt eo e1 m1 e2 m2 ICICI 841k eTe a n CE
2. 0 97 Cl RREPEIRREING BS vaccagicd scree ates cae a actions 100 2002 2008 GNI Ltd and Human Genome Center Institute of Page 3 of 100 Medical Science The University of Tokyo All rights reserved 1 Introduction Cell Illustrator CI is a software tool that enables biologists to model elucidate and simulate complex biological processes and systems 11 It allows researchers to model metabolic pathways signal transduction cascades gene regulatory pathways as well as dynamic interactions of various biological entities such as genomic DNA mRNA and proteins Cell Illustrator models are used to visualize biopathways interpret experimental data and test hypotheses In addition it provides researchers with model diagrams of publication quality and simulation result charts Cell Illustrator has been successfully utilized to model biological pathways like Circadian Rhythms of Drosophila Glycolytic pathway and Fas ligand induced Apoptosis 1 2 3 4 5 6 and 7 Using the graphical user interface of Cell Illustrator researchers can create and simulate their own molecular pathway models Some of the key features of Cell Illustrator are a Pathway construction and visualization construct models graphically by drag and drop entities that represent bio components in the biological pathways b Pathways simulation specify mathematical formulas for biochemical reactions in the pathway for simulation The simulation can be run inside
3. 4 ghfdt ivsi hhf rsidpg kedt lef ikvvsrheqelpailddtt lsgedrnahl nalkmncevalirllesfetmasgtnlydldlggkekkartkaahef dweeergpi lq ltl laldirhlwnhsi iesefvelyteceyr lenpt inhgknretreaithl ly H altrynhmlsatyvki iqml ghfehlapylyvaayvelwatdyemks iveeivreleqke pgelsrdpesetkefaaf ltelaervpai lmesmci l Idhldgenymmrnaw laaman size total sequence lengtht J urrent_position ml length 1 1 total sequence substring 0 current_position return ml Figure 67 Selected succeeding states of the model in Figure 66 are presented in Figure 68 transcription transcripted_protein_sequence mi mapgmyethlpls transcription transcripted protein_sequence mi mapgmyethipispeellksqqvaqyygeyl Figure 68 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 61 of 100 5 7 Preferences Dialog cy Preferences Entity Figure gt Continuous Name variable alue HProcess igure Continuous Discrete Generic Text Name speed Script alue Delay Script y alue Connector igure Line inh Size 22 22 Inhibitory Fill Col Association E _ None a igure Arrow Line Color E Black z Process inte Stroke f 7 T Association Figure Mame arable HOther Figures Frame Text Area Canvas L Canas Options XML File L Save Options Simulation Animation Options Extra Options Restore Defaults
4. Gene Networks International Cell llustrator User Guide 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved TABLE OF CONTENTS TINTRODUC TION bosctcusestscensivaidevduamucarstueveuumenessvuardiuaeniesetioeuiescbueneiuserainensends 4 2 GETTING STARTED virecan anaana aae EEan 5 2 1 Cell Illustrator Online and Desktop Versions cccccccecceceeeeeeeeeeeeeeeeeeens 5 2 2 Run Cell Illustrator Online CIO cc ceccseeeseeteeeteeeteeeseeesneeeeeeeeseeeeaees 5 2 3 Install and Run Cell Illustrator desktop VErSiON cccccceeeceeeeeeeeeeeeeeeees 6 2 GB ANCEHSING MOnato Nisei a T suidncnunindtedad ante Mudnadesions 8 2 5 Command Line Simulation ENGING cccccceccsseceeeeeeeeeseeeeeeeeeeseeaeeeees 9 26 Cl_ Memory Customization arsena T A 10 3 CELL ILLUSTRATOR ELEMENTS sirain iunea aaan 11 3 GONCEDE OVEIVIEW riie E nada dowediene sees 11 SL ENUI A E 12 DO CONMCCION S manian Ea E E E G 13 SA POCE SSES a e E a a a ee e eee Per ee 14 3 0 Element ComDNaUONS ernan a a a a a a 15 3 6 Discrete Elements Example 1 ccccceccecceeeeeeeeeeceeceeeeeeeneeneeeeseeeaeeseees 17 3 7 Discrete Elements Example 2 Connector Speed ccccecceeeeeee enna es 18 3 8 Continuous Elements Example 1 cccccecceccseeeeeeseeteeeeeeeneeeneeeeeeeeaeees 19 3 9 Continuous Elements Example 2 cccccccecceceeseeeeeeeete
5. Institute of Medical Science The University of Tokyo All rights reserved Page 91 of 100 Fas iger cytochrome c Fas ligand trimer Pores a Fas ligand trimer Fae receptor complex a sh spase 9 e I 3 i DFF mie O Cleaved DFF45 c DFF45 cleave 7 m _ object of mi tochondri on release cytochrome c from mitochondrion Figure 2 A CI model representing the Fas induced apoptosis obtained from Figure 1 For Bid m11 Procaspase 9 m21 Procaspase 3 m25 DFF m30 DNA m37 the initial concentration of each compound is assumed to be 100 On the other hand for FADD m4 Procaspase 8 m5 Apaf 1 m17 dATP ADP m18 when two compounds react together without the stimulation of apoptosis the initial concentrations and the rate are assumed to be 39 039 and m1 m2 5000 respectively to keep the stable state condition Each compound is assumed to be produced by the rate of 0 5 represented by a process without any incoming connector and to degrade by the rate of its concentration divided by 200 represented by a process without any 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 92 of 100 outgoing connector which keeps its concentration at 100 under the stable state condition This degradation rate also applies to other compounds in the network The rate of other processes is determined roughly by following T
6. Table 3 Input Connectors d t v t Process is enabled at intervals of the time d t while m t gt w t holds Om w t me c t Process is enabled at intervals of the time d t while m t gt w t holds but no quantity change m t w t M t Process IS not enabled while m t gt w t holds Not Allowed Om w t a Process is enabled while m t gt w t holds but no quantity change m t w t C Process IS enabled m t gt w t holds not while 3 6 Discrete Elements Example 1 m t w t d t v t Process is enabled at intervals of the time d t while m t gt w t holds Om w t mm c t Process is enabled at intervals of the time d t while m t gt w t holds but no quantity change a ci Process is enabled m t gt w t holds not while iT bo F a am Figure 4 v t Process is enabled at the speed of v t while m t gt w t holds is enabled while m t gt w t holds but no quantity change m t w t C Process IS enabled m t gt w t holds Process not while me The model in Figure 4 consists of two discrete entities el and e2 and the discrete process p1 between them The entity el is connected to p1 with the input process connector cl The process p1 is connected to e2 with the output connector c2 The kinetic style of the process is Custom The entities el and e2 have the initial value 10 and 0 respectively Th
7. Behavior of Translocation Process Many processes in a cell show stochastic behavior In order to model a stochastic translocation process the model in Section 4 2 was modified by applying a script in Figure 64 to the Activity option of process p1 The results of the simulation of the discrete model with the sampling interval equal to 1 are presented in Figure 65 The figures compares 2 discrete processes one randomly activated and the second always active process Note that the value changes of the p53_ nuclei random entity are not steady chart 2 xj Copy To Clipboard p53 nuclei ac v p53_cytoplasm E AL Bag p53 nuclei fran pS3_cytoplasm i Dj mA oO fh O th th a a a ee ee ee a ee Figure 65 5 6 1 2 Example 2 Transcription Process with Generic Entities and Processes Generic entities are of two value types For example you can use an entity that holds a String variable to describe a translation transcription process with a generic process transcription transcripted protein sequence mi Figure 66 In this example a script shown in Figure 67 is assigned to the Updater property of the connector 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 60 of 100 Script Editor File Variable transcripted protein otal sequence mapamyefhlplspeel lkegevnqyyygevlsikhl ppl raf gaat ragep lam
8. F AmI pha mi6 2 p epg P33 mMdmz mig 100 at 1 0 0 0 cli 0 0 mem mL 50 ps3 mig c19 p7 EH po3aomdm2 _ m21 100 cit 1 0 0 0 c20 0 0 mem moe 50 0 p53 m22 099 pe cod p53 _medm2 m4 100 0 ao 1 0 0 0 c23 0 0 mdm2 _ m24 5o p53 m25 c25 pg c27 p53 _medm2 m27 100 0 i 1 0 0 0 c26 0 0 mdm m26 50 0 Figure 28 The simulation results of the continuous model are presented below Figure 29 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 30 of 100 Denaren x Copy To Clipboard po pane C ne cr T psa mdn2 ba Figure 29 If the complex speed strictly depends on the value of p53 m1 and mdm2 m2 the speed of the complex process needs to be m1 m2 400 the factor 1 400 is an example for the model where all elements are of the continuous type as in Figure 30 pos mi pi c3 pos meme ms 100 0 mi m2v400 mem iY 50 0 Figure 30 The results of the simulation are presented below Figure 31 Denaren x Copy To Clipboard Figure 31 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved P age 31 of 100 4 5 Tetramerization In a cell some proteins create a self complex under certain conditions For example p53 can create a homotetramer The process of creating tetramers is called tetramerization The th
9. Remote simulation finished is displayed to the user In this dialog box you are prompted whether to open the log with CI Player 6 6 Customized Simulation Engine CI Server uses the standard simulation engine supplied with Cell Ilustrator However an advanced CI user who would like to integrate his own simulation program into CI can do so by replacing the default engine at the server with his own program The user engine must fulfill the following requirements Read the input file from a CI model file 1 e the model must be in the CSML format Write the simulation log file in the CSML format to the output directory Write the simulation status file in a format compatible with the CI Server Check for the stop file optionally For more information see the Cell Illustrator Server Installation Manual 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 77 of 100 7 Tools for Model Analysis 7 1 Navigator For a large model the Navigator dialog Figure 82 provides a convenient way to locate and display in the canvas window a desired part of the model The Navigator displays a small view of the entire model with a green rectangle marking the portion of the model currently shown in the active canvas You can drag this rectangle around in order to navigate the model in the canvas window cel ilustrator 2 0 R Fie Edit Element Simulation wiew Dial
10. Toggle Simulation History Dialog on off open or close the Simulation History dialog The dialog contains the list of all simulation log files for the active model file see Section 6 3 2 A Toggle Edit Units Visible Status open or close Edit Units dialog The dialog provides defining and changing units G Toggle Biological Elements Dialog on off open or close the Biological Elements dialog The dialog contains customized entities processes and images for modeling biopathways see Section 5 4 4 e e Toggle Graph Layout Visible Status open or close Graph Layout dialog to set elements on the canvas in the specified way Toggle Finder Visible Status open or close the Finder dialog to search for elements in the active canvas see Section 7 2 k Toggle Pathway Search Visible Status open or close the Pathway Search to search for elements in the active canvas see Section 12 Open Show All Dialogs open all dialogs X Close All Dialogs close all dialogs 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 46 of 100 5 4 Element Settings Dialog and Element Lists Dialog The dialogs that show the details about the variables and equations in the model are the Element Lists see Figure 53 Figure 55 and Figure 57 and the Element Settings see Figure 54 Figure 56 and Figure 58
11. Type Kinetic Script Kinetic Value Firing Style Firing Script and Visible properties are displayed Figure 57 When a single connector is selected in a canvas the Element Settings dialog displays its properties on two sheets the Connector sheet and the Figure Sheet Figure 58 The major properties that you can edit in the Connector Sheet are Type Connector Firing Style Firing Threshold and Firing Rule A connector can be of the 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 52 of 100 process inhibitory or association type The Connector Firing Style option defines how the condition for activating the connector is calculated It should not be mistaken with the process Firing Style parameter The Connector Firing Style can take the Threshold Rule or No Check value In the Threshold style the value of the Firing Threshold determines if the connector is enabled if the current value of the source entity is larger than the threshold then the connector is enabled In case the Connector Firing Style is set to Rule the behavior of the connector is defined by a boolean expression script in the Firing Rule field No Check option always enables the connector Mame ez Type inhibitory Firing Style Firing Threshold Firing Rule Show Element Show Mame how Firing Operation Show Firing Style References Comments Figure 58 5 4 4 Biolo
12. University of Tokyo All rights reserved Page 98 of 100 100 00 p Maho oo m1 0 00 Figure 5 Sampling interval is 10 pt 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 99 of 100 10 11 12 13 CI References Doi A Fujita S Matsuno H Nagasaki M Miyano S 2004 Constructing biological pathway models with hybrid functional Petri nets In Silico Biology 4 3 271 291 Doi A Nagasaki M Fujita S Matsuno H Miyano S 2004 Genomic Object Net II Modeling biopathways by hybrid functional Petri net with extension Applied Bioinformatics 2 185 188 Matsuno H Doi A Hirata Y Miyano S 2001 XML documentation of biopathways and their simulations in Genomic Object Net Genome Informatics 12 54 62 Matsuno H Doi A Nagasaki M Miyano S 2000 Hybrid Petri net representation of gene regulatory network Pacific Symposium on Biocomputing 5 341 352 Matsuno H Fujita S Doi A Nagasaki M Miyano S 2003 Towards biopathway modeling and simulation In 24th International Conference on Applications and Theory of Petri Nets CATPN 2003 Volume 2679 Lecture Notes in Computer Science pp 3 22 Matsuno H Murakami R Yamane R Yamasaki N Fujita S Yoshimori H Miyano S 2003 Boundary formation by notch signaling in Drosophila multicellular systems experimental observations and gene network modeling by Genomic Object Net Pacific Symposium on Biocompu
13. dATP and procaspase 9 and activates caspase 9 The caspase 9 cleaves procaspase 3 and activates caspase 3 The other pathway occurs when caspase 8 cleaves procaspase3 directly and activates it The caspase 3 cleaves DNA fragmentation factor DFF 45 in a heterodimeric factor of DFF40 and DFF45 Cleaved DFF45 dissociates from DFF40 inducing oligomerization of DFF40 that has DNase activity The active DFF40 oligomer causes the internucleosomal DNA fragmentation which is an apoptotic hallmark indicative of chromatin condensation Note this is not a CI model The pathways consist of several steps lt where two different pathways from caspase 8 are assumed gt and many molecules including Fas receptors the caspase family which includes aspartic acid dependent cysteine proteases and products of their zymogens the Bcl 2 family which includes pro and anti apoptotic proteins cytochrome c and DNA fragmentation factor The apoptosis starts from the Fas ligand binding to Fas receptors and ends in the fragmentation of genomic DNA which is used as a hallmark of apoptosis Thus the amount of DNA fragmentation can be assumed to be proportional to the cell death We have designed a model through utilization of the known facts about the Fas induced apoptosis pathways shown in Figure and biochemical knowledge about reactions Figure 2 shows the model representation that we have described with Cell Illustrator 2002 2008 GNI Ltd and Human Genome Center
14. its biological meaning and you can also enter the initial value of the entity The current value of the entity is displayed once simulation data is available To include the simulation data for this entity in a simulation output file check the Log box To make an entity public for inclusion in higher level models see Section 8 2 check the box in the Public column You can also control if the entity should be visible or hidden in the canvas by setting its corresponding checkbox in the visible column 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 47 of 100 rial Vale A ca y ble me 100 100 Se ee E E a Show Options Figure 53 If an entity is selected in the canvas and the Element Settings dialog is activated the properties of the selected entity are presented in the Entity Sheet and the Figure Sheet Figure 54 The type of entity cannot be edited in Element Lists However it can be changed in the Element Settings dialog A continuous entity can only have the Double value type while a discrete entity be it Integer or Long or a generic entity can have the value type String or Boolean The Maximum Value and Minimum Value fields define the allowable value range The display properties associated with the selected entity can be modified in the Figure Sheet 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University o
15. local simulations are interactive 1 e the canvas 1s updated during the simulation run and you may pause or move through the simulation However you can neither edit another model nor run another simulation in this mode Thus running simulations in your CI Workspace is recommended for debugging purposes and simulations that are not time consuming For large models and lengthy simulations sending a simulation job to a remote server is recommended The Cell Illustrator Server CI Server is a companion product of Cell Illustrator and it is not a part of the standard CI package Hence this option 1s available only to users who have additionally purchased the CI Server One server can be accessed by multiple users of the Cell Illustrator desktop package or alternatively one can connect to more than one CI Server from your desktop The remote server mode of running simulations is also useful 1f you intend to run many simulations in parallel The server mode of simulations does not drain upon the resources of your computer so you may continue to work with CI or other applications after submitting many simulations to a server Technically CI Server is a web service installed at an Apache Tomcat web server A remote simulation is completed in three steps 1 upload the model input file to the server 2 execute the simulation using the engine installed at the server and 3 download the output file Simulation Log The CI Server installation procedu
16. net time pt The Simulation Time option sets the duration of the simulation process in Petri net time If the Simulation Time is set 2000 as in Figure 74 the simulation lasts from O pt to 2000 pt Note that the value of the Simulation Time parameter does not necessarily correspond to the real time it takes to execute the simulation on your machine The latter depends also on the complexity of the model additional simulation parameters such as the Sampling Interval and the hardware on which the Cell Illustrator runs The Sampling Interval specifies the interval at which simulation steps are executed Again the Sampling Interval is defined in Petri net time units At each step model state changes are computed If the interval is 0 1pt as in Figure 74 the model changes its state at 0 1 pt 0 2 pt 0 3 pt 2000 pt If the Sampling Interval becomes smaller the simulation becomes more accurate but calculation performance decreases For example for 1 pt simulation Cell Illustrator needs to perform 100000 simulation steps if the interval is 0 00001 pt and only 10 steps if the interval is 0 1 pt The minimum sampling interval allowable in Cell Illustrator is 0 000001 pt You should be aware however that simulation with the minimum sampling interval may take a very long time If the sampling interval becomes very small the simulation result will be similar to a differential equation model see Appendix B The Weak Firing option solv
17. of the inhibitory connector 4 10 Catalysis Enzymes catalyze a variety of chemical reactions The model of an enzymatic reaction can be created with the following steps The first three steps are similar to the translocation process in Section 4 2 Step1 Create two entities and change their names e g p53 and p53 P and set their values e g 100 and 0 Step2 Create a process and set its name e g catalysis Connect p53 to the process and connect the process to p53 P Step3 Set the catalysis speed by editing the speed option of the catalysis process the default speed is 1 0 Step4 Create one entity and change its name e g CAK Step5 Create a process and connect it to the CAK entity with the process connector Step6 Connect the CAK entity to the catalysis process with the association connector Step7 Change the threshold of the association connector to 1 the default value is 0 The type of CAK entity is continuous With the model in Figure 46 one can simulate the situation where CAK enzyme concentration is growing If the association connector threshold is set to 1 the catalysis process will be enabled when the CAK entity value reaches 1 level 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 39 of 100 p2 CAK m3 1 0 a0 i i 110 i p53 _ mi catalysis pos P ome O gt r4 ae O 10 0 1 0 0 0 Figure 46 The
18. offers a method for performing a serious of simulations in a bulk on the same model with varying the selected parameter value s This mode allows comparing the simulation results when initial parameters differ It allows 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 73 of 100 checking for example how the amount of control entity influences the system or choosing the right value of some kinetic constant The changing parameters are selected and the range of their values is defined in the SECG Passing Parameter dialog box After that for each combination of parameter values a simulation is run This mode is faster than running each simulation as a different SECG simulation and it is run outside the Cell Illustrator SECG Passing Parameters Variable Used mi OO Eo The results of Passing Parameter simulation are presented as 2D or 3D charts After running Passing Parameter simulation a control panel window appears which allows from comparing simulation results by selecting the chart type and parameters values The charts always present how the entities depend on one selected entity while the rest of the parameter values is fixed Parameterized Simulation Result 1 a Ed ariable entity 2D chart 7 6 4 2 Advanced SECG topics The SECG program is composed of two components one holds common functionality the second contains information specific
19. results of the simulation are presented below Figure 47 x Copy To Clipboard Figure 47 If the threshold of the association connector is modified to 5 Figure 48 the results will be different The results of the simulation are presented below Figure 49 p2 CAK mo 1 0 0 0 i i i i i 510 i i p53 ma catalysis po3 P _ m5 a O iO 10 0 1 0 0 0 Figure 48 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved P age 40 of 100 x Copy To Clipboard Figure 49 In the presented reaction CAK influence on the model is enabling disabling the reaction The CAK activity will be adjusted with the threshold of the association connector However in most of the systems the reaction rate depends on enzyme concentration and substrate concentration In the example in Figure 50 the speed of the catalysis process is defined with the equation m1 m2 10 The model is made of continuous type elements p2 enzyme ma i 1 0 0 0 i i 0 0 i substrate _ mi pi l product m2 Tee 10 0 i m1 ma 0 0 0 Figure 50 The results of the simulation are presented below Figure 51 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 41 of 100 Chart 2 enzyme E substrate E product Figure 51 2002 2008 GNI Ltd and Human Genome Center I
20. the workspace window in an interactive mode or sent to a remote server Cell Illustrator Server c Simulation replay the simulation results can be logged and then visualized and analyzed in Cell Illustrator Player d Simulation animation Cell llustrator s simulation results can be edited into an animation script file and then be used as the input to Cell Animator a module that produces high quality presentation of the studied mechanism s dynamic behavior e Model import previously created models or sub models can be imported from SBML CellML files or from public or proprietary libraries and databases like KEGG 9 12 The documentation for the Cell Illustrator software includes the manuals listed below Cell Illustrator User Manual Introduction to Cell Illustrator system Cell Illustrator Reference Manual Detailed description of Cell Illustrator functionalities Cell Illustrator Player Reference Manual Detailed description of Cell Illustrator Player functionalities Cell Illustrator Server Installation Manual Installation instructions for Cell Illustrator Server product for IT System Administrators This manual describes version 4 0 of Cell Illustrator Online CIO4 0 and version 3 0 of Cell Illustrator desktop CI3 0 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 4 of 100 2 Getting Started 2 1 Cell Illustrator Online and Desktop Ve
21. the detailed mechanisms of the selections of these two pathways from caspase 8 are still unclear and need to be studied in future Since the presence of autocatalytic process is proposed in caspases 2 it is included in our model Figure 4 which increases the DNA fragmentation as shown in Figure 3 pro caspase Figure 4 A model representation of autocatalytic process in Figure 2 However if the large rate of the autocatalytic process 1s assumed in the caspase reaction the DNA fragmentation becomes independent of the Fas ligand concentration which then disagrees with the experimental results Therefore we can guess that autocatalytic processes must be slow if they are present To examine the effect of autocatalytic processes of caspases on the apoptosis by Fas ligand DNA fragmentation is simulated when the stimulation by Fas ligand 1s stopped after a short period Table 2 shows a simulation result where the apoptosis proceeds more with the increase of the autocatalytic rate of caspases even for a short period stimulation 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 94 of 100 Table 2 DNA fragmentation at four autocatalytic rates of caspases rateO 0 ratel mA mB 80000 rate2 mA mB 40000 and rate3 mA mB 25000 which are assigned to the process T4 in Figure 4 The stop time represents the period after that Fas ligand stimulation is stopped The initi
22. with CI The examples have been designed to help the novice user to become familiar with the Hybrid Petri Net modeling method and notation 4 1 Degradation In a cell nearly all mRNAs and proteins fragment into parts at certain speeds This fragmentation activity is called degradation A model of degradation can be created with the following steps Step1 Create an entity and change its name e g p53 and set its value e g 100 Step2 Create a process and set its name e g degradation Step3 Connect the p53 entity to the degradation process with the process connector Step4 Set the degradation speed by editing the speed parameter of the degradation process the default speed is 1 0 In Step1 and Step2 a discrete entity and a discrete process a continuous entity and a continuous process a continuous entity and discrete process combinations are possible Figure 16 g mi p m3 10 C threghald ae and Figure 16 The results of the simulation are presented below Figure 17 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 24 of 100 a Chart 2 xj Rasterize Copy To Clipboard Figure 17 If the degradation speed depends on the value of p53 protein say m in Step 4 the speed should be made dependent on m Figure 18 e4 m4 e5 md eG _ m degradation degradation degradation 1 migi g moi 0 m
23. 0 1000 2000 3000 0 1000 DFF40 oligomer DNA fragment 0 1200 90 as 00 600 30 300 t t t 0 0 2000 3000 0 1000 2000 3000 1000 2000 3000 90 1200 900 60 600 30 30 300 t t t 0 0 2000 3000 0 1000 2000 3000 1000 2000 3000 90 1200 900 60 600 30 30 300 t t t 0 0 2000 3000 1000 2000 3000 1000 2000 3000 Figure 5 Simulated time courses of some intermediates during apoptosis for the Fas ligand concentration n 210 450 and 600 References 1 Harada H 1999 Regulation of apoptosis by BH3 domain only proteins Jikkenigaku 17 1603 1606 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 95 of 100 2 Hugunin M Quintal L J Mankovich J A and Ghayur T 1996 Protease activity of in vitro transcribed and translated Caenorhabditis elegans cell death gene ced 3 product Journal of Biological Chemistry 271 3517 3522 3 Kuwana T Smith J J Muzio M Dixit V Newmeyer D D and Kornbluth S 1998 Apoptosis induction by caspase 8 is amplified through the mitochondrial release of cytochrome c Journal of Biological Chemistry 273 16589 16594 4 Nijhawan D Honarpour N and Wang X 2000 Apoptosis in neural development and disease Annual Reviews of Neuroscience 23 73 87 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 96 of
24. 1 In this model process p1 continuously transports entity e1 to e2 at the rate of 1 20 of el s concentration m1 Since pl is connected to el with quantity represented by m1 and its output is connected to e2 with quantity m2 the equations for quantities change are dml _ dm dt dt 0 05 ml The value of entity el 1 e m1 decreases while the value of entity e2 1 e m2 increases during simulation The equation 0 05 m1 of process pl governs this change of value per unit time see Section 6 1 for Simulation Details The results of the simulation are presented below Figure 13 x Copy To Clipboard Figure 13 If consumption and generation speed should be different the Connector Custom kinetic style should be used and the corresponding speed add parameters for connectors cl and c2 should be set In Figure 14 the speed of cl and c2 1s 0 05 m1 and 0 1 m1 respectively Figure 14 The results of the simulation are presented below Figure 15 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 22 of 100 chart ERs copy To Cipboard Figure 15 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 23 of 100 4 Biological Process Modeling with Cell Illustrator This section briefly explains how to simulate major biological processes
25. 1 1 0 i pad Ele mg cd translocation c3 P33 cytoplasm _ m10 100 0 0 0 1 0 o0 Figure 20 The simulation results of the discrete model are presented below Figure 21 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 26 of 100 x Copy To Clipboard wane cater p53_nucei a p53_cytoplesm Ml Figure 21 If the translocation speed depends on the quantity of p53 nuclei say m1 as in the degradation example Section 4 1 the process speed should be made dependent on m1 10 in Step3 The factor of 1 10 was used in the example in Figure 22 p a nuclei _ m1 ef translocation e3 PaS_cytoplasm _ mz e e 100 0 i mio 0 0 Figure 22 The results of the simulation are presented below Figure 23 x Copy To Clipboard tame coer p53_nucei am p53_cytoplesm T Mi Figure 23 4 3 Transcription Transcription modeling can easily be done using CI A model can be created with the following steps 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 27 of 100 Step1 Create a process and set its name e g transcription Step2 Create an entity and change its name e g mRNA p53 Step3 Connect the transcription process to mRNA p53 entity with the process connector Step4 Set the transcription speed by editing the Speed Add option of the transcri
26. 100 Appendix B Using Differential Equations in Cl In this example the following differential equations are modeled with Cell Illustrator 1 p rt mit ml 0 mdt mt 5 ml t dt m1 0 10 m2 0 0 ooog i 2 00 Figure Differential equation results mi mi me me om ai M c2 10 0 0 0 miji 0 a0 Figure 2 Figure 1 shows the differential equation results For the model in Figure 2 the simulation results for runs with Sampling Interval equal to 0 1 pt 1 pt and 10 pt are shown in Figure 3 Figure 4 and Figure 5 respectively Comparing simulation results to those in Figure 1 you can easily notice that the smaller the sampling interval becomes the more similar the simulation result is Note that Cell Illustrator offers modeling features that are alternative to modeling with differential equations and allow the simulation of complicated systems in an easier way 1 discrete behavior and continuous behavior can be mixed in one model in Cell Illustrator 11 the notion of delay can apply to discrete processes 111 activity and threshold concepts can be used in CI 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 97 of 100 sot Aalst o E m Figure 3 Sampling interval is 0 1 pt oog m ee 00 Figure 4 Sampling interval is 1 pt 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The
27. 4 6 2 How to Set Up and Run Simulaties eeina aa 66 6 3 Viewing and Analyzing Simulation ReSUItS ccccececeeeeeeeeeeeeseeeeaeeees 71 6 A4 RUNNING SECS SIMULIUM errs E E 73 6 5 Remote SIMUlAtiION Servers iinitan a a i 15 6 6 Customized Simulation ENGINE ccsa SE 77 T FOOLS FOR MODEL ANALY SIS irni a 78 if A INQVIG QUOR ssasntecticdutdorerutesuiootiueteduivediastdes E EEEE 78 TEPA WANN CRG I sine ett cerca tcc ace cates a E tee al A nA OT TERE ee 19 ko MIACANG PI ALCS unarne N E 80 R EXPO ODIOM S rroi aa E AA haa Goad rarmntaten 81 8 MAKING COMPLEX MODE LG ccccccccceeceeeeeeneeeeeeeeeenseeeeseneseneseneseaes 82 8 1 Grouping and UNGrOuPIN ccccceccceeecseeeceeeeeeeeceeeceeeeseeeseeesseeeseeeseneeas 82 SZ MASLALCIIZAULON oa E a A 82 7 OSGi EIS a a a I ne eae naaiaee iaaeeas 84 oA BOPA C Sracistiatoit enero Retearoeid tieenaind ed aguas GE E alae aio ton bee caial ge pane eee Staaten 85 8 5 Import of CelIML or SBML MOEIS cccecceecceeeeeeeeeeeeeeseeeeeesaeeeeeeeaees 85 8 6 Import of CSO and BIOPAX MOGEIS cccccceccceeceeeeeeeeceeeseeeaeeeeneeeeeeeess 86 9 CELL ILLUSTRATOR PLAY ER ivvcesinsitviccessesticeti cecabaiineivecicevseuiintsnequaetateann 87 10 CELL ANIA TOR eae inci ietceciiensnncitenvevessccve saacatwacaus aan cudnsmadastaeceaceuaeeabacanect 89 APPENDIX A TUTORIAL MODELING FAS LIGAND INDUCED APOPTOSIG 90 APPENDIX B USING DIFFERENTIAL EQUATIONS IN Cl
28. CRI ica Aa T E aa on View Dialog Window Help i La Mk OODE LL i FERRES S ER H e C R New Canvas 2 B Ne Pa a a EA a Lo a T Gii v o 5 Dl a a zl a Ea 2 E e 2 hed k K S 2 I be T ii m lt lt O aa Gp O ca cm a El X al 3 gt i Pl oo gt Dd rst o a o il ul a ul gt gt gt gt gt G G S e e i j P L L e e Figure 88 Edit Parts usage 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 84 of 100 8 4 BioPACS BioPACS is a Web system that converts metabolic pathway maps from the KEGG Web site to the XML file format used by CI 9 12 To connect to BioPACS select the option File Import BioPACS or click on the icon in the Edit toolbar This opens the KEGG BioPACS dialog Select a KEGG map name and click on the Download button to convert it with the BioPACS tool and open in Cell Illustrator as a new canvas Figure 89 KEGG BioPACS Metabolic Pathways 14 Carbohydrate Metabolism slhycols ion f n nesis Edit Element Simulation View Dialog Window Help N OtrltN H W Oa Ht a feed 11113 flat Ql wi HSEEE a Ctr cent File Ctrl ay SIS VIUCONCOMEHeSIS Citrate c
29. Figure 69 With the Preferences dialog you can set the global properties of Cell Illustrator This section briefly introduces the major options available in the dialog The dialog displays a tree of option types and for a selected part of the tree the list of available options and their settings For example Figure 69 shows the global settings of the figure to be used for display of continuous connectors One can define the size of the figure 1ts color and other parameters The Canvas options define the way the canvas looks The background color and the grid size can be defined here Alternatively you can update major canvas options in the View Settings dialog 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved P age 62 of 100 The XML File Save Options define properties of XML files that store Cell Illustrator models such as the encoding type of the XML format The Simulation Animation Options define visualization parameters of simulation On the other hand the Simulation Extra Options define the default parameters of a simulation process The values of Sampling Interval Plot Update Interval Log Update Interval Max Speed and Simulation Time displayed in the Preferences are used as the default parameters in the Simulation Settings dialog 5 8 Graph Layout Dialog In this dialog box you can beautify the graph layout of the model This function can be e
30. I offers new license types Pro Draw and Classroom The main differences between the license types are in the simulation capabilities of CI and are briefly summarized in the table below CI Pro Full version Includes all simulation and editing drawineg CI Classroom Limited version Simulation of small models only up to 30 entities All editing drawing capabilities CI Draw Limited version No simulation capabilities Full editing drawing capabilities Note This manual describes all functions available in the CI CI Pro license type A given function described in this manual might be disabled or not available in the limited versions CI Draw and CI Classroom 2 5 Command Line Simulation Engine Note this feature is available in desktop version only Cell Illustrator can be used to run simulations in the command line mode This simulation engine works with Cell System Markup Language CSML input output and can be called from user scripts or integrated with a custom GUI Alternatively you can utilize Cell Illustrator Server web service in your modeling environment In both cases you need to obtain a CI Server license The example script C Engine to run CI simulation engine in the command line mode can be found in the CI installation folder The example script contains the java command which starts the simulation engine in the form JavaExecutable classPath mainClass inputFilePath 2002 2008 GNI Ltd and Human Genome Center I
31. Ka EH e H el Print Area C Whole Graph visible 4rea Selected Rectangle x From fa pas mani ut z kard 7 na s and Gaz Print Area Selection A Mark the area to print with the mouse and click OK RRO ee SR a Lae X 0 Print the canvas total free 98MB 29MB Figure 84 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 80 of 100 7 4 Export Options Analysis of model structure and simulation results can be also carried out outside Cell Illustrator with third party tools The options Export Entity List Export Process List and Export Connector List in the File menu store respective data in a file in the CSV Comma Separated Values format Such a file can be later uploaded to a spreadsheet for example The option Export CSO Model available since version 4 0 allows to export model to Cell System Ontology 3 0 format Cell Illustrator 2 0 le x File Edit Element Simulation Yiew Dialog Window Help im Aele z iai w D LlLl LARS Clee i A E ee H a pen r x Open Recent Files gt 2121 x Close File Ctrl W a Close All Files Ctrl Shift W d F Save Ctrl S El Save As Ctrl Shift S fooga aapa CyN pet TID bee oy Import Process List lt 7 Look amp Feel a comsen m20 EJ Microsoft Excel e
32. They can be opened either by selecting the options Element Settings and Element Lists respectively in Dialog menu or by clicking on the buttons andil in the Property Dialog toolbar The Element Lists dialog consists of three sheets the Entity Sheet the Process Sheet and the Connector Sheet Each sheet contains a table of all of the entities processes and connectors respectively in the active canvas and their main properties You can easily change a property of an element by double clicking on the corresponding cell in the table Note that the columns which can be edited have a white background while the columns that cannot have a gray background To change an element property displayed in a gray out column you need to select a corresponding element in the canvas and open the Element Settings dialog The Element Settings dialog displays all the properties of a single element selected in the canvas The dialog may display depending on the type of selected element one or two tabs from the following set Entity tab Process tab Connector tab Figure tab and Image tab Any modifications made in the Element Settings will be immediately reflected in the Element Lists and the canvas and vice versa 5 4 1 Entities For entities several properties are available for editing in both dialogs see Figure 53 and Figure 54 A default setting 1s automatically displayed if you have not entered any value You can change the name of the entity to reflect
33. able 1 Synthesis and catabolism processes are added in the model for all proteins Autocatalytic processes are also added in the model to all caspases since they exist as proenzymes The pathway from caspase 8 to caspase 3 1s assumed when the caspase 8 concentration 1s over 30 Protease is often synthesized as a proenzyme zymogen and changed to active form by other enzymes or by itself So the autocatalytic process is added to every caspase reaction Table 1 Functions assigned to continuous processes in the simulation of apoptosis induced by Fas ligand where mA and mB represent the contents of the corresponding continuous entities Rate Unimolecular reaction Bimolecular reaction Self effacement mA 200 Oligomer mA 20 mA mB 10000 Monomer mA 10 mA mB 5000 Enzyme binding mA 5 mA mB 2500 Enzyme reaction mA 10 By using the apoptosis scheme modeled as a model we simulated the amount of DNA fragmentation by varying the Fas ligand concentration and Figure 3 shows the simulated relationship It shows that under very weak stimulation very low amount of Fas ligand DNA fragmentation does not occur since the stimulation stops at the intermediate point because of the assumption of degradation processes With the increase of the stimulation the reaction proceeds to the backward intermediates and DNA fragmentation cell death occurs finally which increases with the increase of the Fas li
34. al Fas ligand concentration is set to be n 600 Variables mA and mB represent the contents of the continuous entities going into Ta DNA Fragmentation Stop time rate0 ratel rate2 rate3 10 0 0 0 1169 15 251 442 746 1862 20 417 581 885 2048 Figure 5 shows simulated time courses of the model in Figure 2 with CI Some intermediates during apoptosis at three levels of Fas ligand concentrations are measured These time courses might be useful to plan new experiments such as addition of inhibitors to various steps However it is necessary to estimate the realistic rates of each reaction by comparison with the experimental data It is also necessary to add other pathways through the Bcl 2 family 1 or p53 to describe the real apoptosis held in various cells and by various inducers CI is a very useful tool for biochemists to describe the complex biological pathways semi quantitatively on a figure n 210 Fas ligand caspase8 cytochrome c 600 40 60 20 30 15 300 20 30 10 10 t 5 0 0 0 0 0 1 2 3 0 1000 2000 3000 0 1000 2000 3000 n 450 600 40 60 20 30 15 300 20 30 10 10 t t 5 o 0 o 0 0 1 2 3 0 1000 2000 3000 0 1000 2000 3000 n 600 600 40 60 20 30 15 300 20 30 10 t 10 t t 5 0 0 0 0 0 1 2 3 1000 2000 3000 0 1000 2000 3000 0 caspase9 caspase3 90 60 30 t 0 1000 2000 3000 0 1000 1000 2000 90 60 30 t 0 3000 0 1000 90 60 30 t
35. application for replaying and detailed analysis Cell Illustrator Player 1 6 File Replay View Dialog Window Help s n TMI i e cil ad OO lt l 39 1269 Fas aN Lii z j Fas ven ene Fas receptor complex nse a 77 Apoptosis JL ee caspase 1370 00 pt caspase9 4 39 caspasea E ooo h 370 0 Current Time 1370 0 Total Time 2000 0 Current Step 1371Total Step Count 20014 totalfree 12 7MEV6SMB Figure 90 CI Player Workspace and Charts The main window of CI Player very much resembles the Workspace window in Cell Illustrator However instead of the Simulation menu and toolbar of Cell Illustrator CI Player has the Replay menu and toolbar with commands for moving forward and backward over the time series in the log The model state entity values and process status at a given simulation time point is visualized in the CI Player canvas window On the other hand charts in CI Player show entity value changes over the whole simulation period with a vertical bar indicating the time point visualized in the canvas Figure 91 Process state indication in CI Player from left to right firing waiting and not firing 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 87 of 100 One advantage of replaying the simulation from a log as opposed to rerunning it in Cell Illustrator is the capability to deterministically repeat the simul
36. ation even for a model of stochastic nature Also you can quickly navigate to a specific part of simulation and step through it forwards and backwards for debugging purposes For more details please refer to the CI Player online help and CI Player Reference Manual You can open them from the CI Player Help menu 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 88 of 100 10 Cell Animator The Cell Animator provides a platform to animate the biological pathways modeled with the Cell Illustrator Please refer to Cell Animator User Manual for details 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 89 of 100 Appendix A Tutorial Modeling Fas Ligand Induced Apoptosis Apoptosis Induced by Fas Ligand Apoptosis programmed cell death is known to participate in various biological processes such as development maintenance of tissue homeostasis and elimination of cancer cells Malfunctions of apoptosis have been implicated in many forms of human diseases such as neurodegenerative diseases AIDS and ischemic stroke Reportedly apoptosis 1s caused by various inducers such as chemical compounds proteins or removal of NGF The biochemical pathways of apoptosis are complex and depend on both the cells and the inducers Fas induced apoptosis has been studied in detail and its mechanism has b
37. by the formula x Simulation L Log Options focess hhikbitory Association arme Threshold Script alue Object pem Author ame e ore Precision l E 4 Other Figures Economy Log lor di Frame l Ted irca Max Log Steps in Memory hoooo ahas canvas Options AML File cave Options Simulation Animation Options xtra Options og Options Y Restore Detaults Figure 76 There are several log options that you can set in the Preferences dialog In the Simulation Log Options panel you can specify the following parameters Author the name of the person who runs the simulation This name appears in the Simulation History dialog Economy Log specifies whether to minimize the size of a CIL file by not writing identical values The recommended state of this value is on It can be turned off only in special cases Precision this value specifies how to round up floating point values The Log Precision defines the minimal difference between two floating point values in the log Max Log Steps in Memory this value specifies the size of the log buffer in the RAM memory If the number of steps in the log exceeds this value the buffer is emptied by writing the log to the 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 69 of 100 hard disk The simulation is suspended during the write to disk operation and resumes w
38. ccount When you register on Cell Illustrator website you will get access only to some of CIO modules To get access to other modules you will need a special license For details please contact GNI or your local Cell Illustrator reseller 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 8 of 100 2 4 2 CI Desktop To run Cell Illustrator you need to obtain a license file from GNI or your local distributor of Cell Illustrator Copy a license file to a directory where the Cell Illustrator is installed For example on Windows systems the Cell Illustrator is installed to C Program Files GNI Cell Illustrator if the default installation settings are used A license can be bound to your machine and or time limited To obtain a hardware bound license you need to send your machine hardware identifier Hardware ID which can be revealed by clicking on the About Cell Illustrator option in the Help menu In the About Cell Illustrator dialog the Hardware ID is displayed as the License Details tab The Hardware ID is also displayed in the Invalid License dialog that appears at the application startup if Cell Illustrator cannot locate a valid license Press the Copy All button to copy the error message and the Hardware ID into the clipboard You can later paste this text in into your message to the Cell Illustrator sales and technical support team Since the version 3 0 C
39. ch version of Cell Illustrator When Cell Illustrator is cited please cite the literature 11 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 100 of 100
40. d CI Non Java bundled Installation To install Cell Illustrator 1 Copy CI3 0x_1 bin to a temporary directory 2 Execute the command chmod x CI3 0x_1 bin 3 Execute CI10x_ bin file To launch Cell Illustrator 1 Go to CI directory 2 Execute CI 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 7 of 100 2 3 5 UNIX 2 3 6 Non Java bundled Installation To install Cell Illustrator 1 Copy CI3 0x_u jar to a temporary directory 2 Execute the command java jar CI3 0x_u jar To launch Cell Illustrator 1 Go to CI folder 2 Execute CI 2 3 7 MacOS X 2 3 8 Non Java bundled Installation To install Cell Illustrator CI 1 Copy CI3 0x_m tgz to a temporary directory 2 Double click on CI3 0x_m tgz 3 Double click on JExpressInstaller 4 A Cell Illustrator icon should appear on the desktop double click to launch Cell Illustrator CI3 0 requires JRE version 1 5 0 or higher The jre1 5 0 for Mac OS X Tiger can be downloaded from http www apple com support downloads java2se50release 1 html You may also choose to install Cell Animator for Mac OS X Go to the directory where you installed Cell Illustrator and double click on the CellAnimator dmg file Then double click on the Cell Animator application icon 2 4 Licensing Information 2 4 1 CI Online CIO To run Cell Illustrator you need to obtain a CI Online a
41. del from a specified file as a singular object that is treated as one not solely for editing purposes but for simulation purposes as well In other words Import As Object allows you to build a hierarchy of models where a high level model contains objects representing lower level models The concept of model hierarchy is demonstrated in the example below 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 82 of 100 Once a model is created in a canvas Figure 87 a and saved in a file it can be loaded Figure 87 b and reused in a new canvas Figure 87 c and Figure 87 d as one element referred to as a hierarchical object In the new canvas processes can access only those entities inside the hierarchical object that have the Public property set to the on value The Public property can be set in the original file in the Element Lists dialog or the Element Settings dialog In the example entities pl and p2 are public Multiple copies of a model can be loaded as objects onto a canvas View Dialog Window Help e View Dialog Window Help _ Pe OOM 1 11 14161812121 Qielw itl 4 Hake el Pe OcOm es 1 11 418 82 Qele iim HESSE New Canvas 2 e 5 53 ii Mose E ke tia fi AUE e fe fev lt ee e1 m1 e2 y m2 CAREM AMARE g E CRR AAAA e
42. e active protein protein active at the speed of phosphorylation which is V t dm0O dt In another words the concentration of protein inactive will decrease by the amount m0 0 1 m1 time and the concentration of protein active will increase by the same amount Once the concentration of protein inactive reaches the threshold value 2 the connector will be deactivated Consequently the process of phosphorylation 1s terminated The end simulation state is displayed in Figure 72 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 64 of 100 proteinfinactive alito phosphorylation nrotein active mi G E C2 O J m1 0 1 m2 g Figure 72 Figure 73 presents a simulation chart for this model di Chart 1 x Copy To Clipboard 100 nave Je AAT a A proteini active AL I W Ja jproteintinactive HE a ral oy 7 Pa o Fi S f Figure 73 When simulating a complex pathway you will often need to construct multiple entities with multiple connectors leading to one or more processes In these cases only when all process and associate input connectors are activated and all inhibitory input connectors are not activated the process is enabled For each simulation step all the activation conditions for all processes are checked and all activated processes are executed in a random order The simulation r
43. e Cell Illustrator adds the notion of continuous and generic processes and quantities which are essential to the description of biological systems 10 The following table presents a summary of the Petri net model elements their explanations and equivalent biological meanings Table 1 HFPNe elements Hybrid Functional Petri Net Elements Petri Net Component Symbol Biological Equivalent Discrete Place A discrete place holds tokens represented as a non negative integer Tokens A token is a unit held in a place Discrete Transition A discrete transition is a discrete process that consumes tokens from their input places and produces tokens into their output places The input and output places are connected to a transition by arcs A transition fires and the process is invoked if all the conditions for firing are met Every transition has a delay time for firing Continuous Place A continuous place holds a non negative real number Marking A variable representing the state of a place 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Discrete Entity A countable biological component or event that is quantified or represented by an integer e g the number of molecules locating at the membrane an event that describes whether a protein is binding to DNA or not Quantity The number of entity items present e g the number of molec
44. e canvas window around the model without using the Navigator dialog or the scroll bars ae Zoom In Zoom in on the canvas e x lt Zoom Out Zoom out from the canvas lt Reset Zoom Restore the default models magnification 2002 2008 GNI Ltd and Human Genome Center Institute of Page 44 of 100 Medical Science The University of Tokyo All rights reserved s Fit In Canvas Restore the default models magnification q Group Group together chosen elements to treat them as one for editing T Ungroup Undo a group operation vA Load Image Load an image and insert the image to the canvas Q Set Color Apply the selected color to selected elements in the canvas as specified in the Select Color Scope e _ A Select Color Tool Select the scope of Set Color ts operation Fill Color Line Color or A Text Color e 4 Set Stroke Change the stroke pattern of selected elements e Toggle Grid Visible Status Turn the background grid on or off Os e gt O Toggle Antialiasing Status Turn on or off antialiasin 2 g g Go to BioPACS URL Import metabolic pathway from public KEGG database see Section 8 4 5 3 Change Element Properties For additional editing of the properties of elements in the model the Property Dialog toolbar at the far right of the window is pr
45. e input connector cl has the Threshold parameter equal 1 The threshold of an input connector controls the activity of process p1 The process pl has the delay parameter equal 1 and the add parameter equal 2 This model can be used for auto catalytic reactions e g protein auto phosphorylation 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 17 of 100 In the HFPNe theory time is measured in virtual time units called Petri net time pt At time 0 the value of el is 10 and the threshold of cl is 2 and the process pl is enabled When all input processes are enabled the connected discrete process is really executed after the delay time of the process In this example the delay parameter of the process p1 is 1 Therefore the process is executed at time 1 Consequently at time 1 the entity el decreases and e2 increases by the speed of process p1 1 e 2 Thus at time 1 the value of entity el and e2 becomes 8 and 2 respectively At time 4 the value of the entity el becomes 2 and the process p1 cannot be enabled because the value of el is less or equal than the threshold of cl At that time the process p1 cannot be enabled and el and e2 reach a stable state The results of the simulation are presented below Figure 5 x Copy To Clipboard Figure 5 3 7 Discrete Elements Example 2 Connector Speed In the previou
46. e rate of entity value changes and interactions among entities Processes are used to model biological reactions such as enzymatic reactions or protein complex formation processes Processes can take multiple inputs e g consume different entities to produce new entities and have multiple outputs e g produce different entities such as breaking down sacharose into glucose and fructose Also like entities processes can be continuous continuous processes discrete discrete processes or generic generic processes depending on the input output and timing Only generic processes are able to handle generic entities ii name migrate phosphorylate translate B iin ste L Jand a Hand delay kinetic script mi 3 i Continuous Discrete Generic Process Process Process gt Figure 3 In general terms if all input connectors of a process are activated the process is said to be enabled see Section 6 1 When a process is enabled the reaction that it represents will fire will be executed immediately or after a certain time interval This will cause a change in value for all connected entities For example if a process represents the reaction of a protein precursor which turns into an activated protein by phosphorylation the decrease of the precursor protein and the increase of the active protein on account of the speed of phosphorylation is simulated once the process fires The exact conditions for enabling a process de
47. e the concentration of doxorubicine is growing If the inhibition connector threshold is set to 0 for continuous processes the translation will be stopped immediately after the first 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 37 of 100 Sampling Interval For a discrete process Translation will be stopped after the first step as during the first step process p1 is enabled p2 doxorubicine m2 p2 doxorubicine m4 1 0 0 00 MRNA pos mi pi MRNA pos m3 0 0 1 10 0 Figure 42 The results of the simulation are presented below Figure 43 Chart 1 Rasterize Copy To Clipboard Wane _ cob doxorobucine di doxorobuce Mal mena pss Figure 43 If the threshold of the inhibitory connector is modified to 5 Figure 44 the results will be different The results of the simulation are presented below Figure 45 p2 doxorubicine _ m2 p2 doxorubicine _ ma 1 1 0 0 MRNA _ p53 mi MRNA pos m3 0 0 1 0 0 Figure 44 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 38 of 100 Chart 1 q x Rasterize Copy To Clipboard doxorobucine m mRNA PSS B dovorobucre E menas E Figure 45 As in Figure 43 and Figure 45 if doxorubicine is treated the transcription of p53 is inhibited The inhibitory activity will be adjusted with the threshold
48. eated here with the following steps Step1 Create three entities and set their names e g p53 mdm2 p53 mdm2 and set their value e g 50 0 50 Step2 Create one process and set its name e g separate Step3 Connect the p53 mdm2 entity to the separate process Connect the separate process to the p53 entity and the mdm2 entity Step4 Set the separate speed by editing the speed add options of the separate process the default value is 1 0 Again 9 combinations of element types in the set p53 mdm2 p53 _ mdm2 separate process are possible in Step and Step2 1 e discrete or continuous discrete or continuous discrete or continuous discrete and continuous continuous continuous continuous as in Figure 38 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 35 of 100 p53 mi gi pt o9h53_mdm2 m3 p53 m16 is pi 2h 53_mdm2 mis 50 f 10 0 0 50 50 f 10 on 50 0 c3 c3 mdm2 _ m2 mdm2 _ m 0 0 p53 _ m4 c1 p1 c2P53_mdm2 _ m6 p53 m19 cf p1 c2P53_Mdm2 _ m21 50 50 T 1 0 ii 50 0 c3 mdm2 mdm2 _ m28 0 0 0 0 p53 _ m7 e p1 c2P53_Mdm2 _ mg p53 m22 cf p1 3P53_Mdm2 _ m24 50 0 mdm2 mdm2 m23 0 0 p53 _ m10 m pt o9P53_mdm2 m12 p53 _ m25 Ei pi c2P53_mdm2 m27 50 0 T 1 0 0 0 50 50 0 10 0 0 50 0 c c mdm2 m mdm2 m26 0 0 0 0 p53 m13 c1 pt oh53_mdm2 m15 0 0 Figure 38 The results of the simulation are presented below Fi
49. ectors An input connector can take all three types In contrast an output connector is always of the process type Correspondingly all types of connectors can connect input entities to a process In contrast only a process connector can connect a process to an output entity E name c2 c3 c4 firimy stre r resh r nocheck threshold rule seript 0 true E gt Process Inhibitory Association a Connector Connector Connector Figure 2 The threshold is an input connector s parameter It can be a value or a script e g m 10 This parameter is used to define the minimum value of the input entities needed for the activation or deactivation of the linking processes A process and an association connector become activated 1f the threshold of the connector is smaller than the value of input entity On the other hand inhibitory connectors are activated when input entity is less than this threshold Inhibitory connectors are used to model repression see Section 4 9 Association connectors are used to model the situation when entities and processes need to be linked together but nothing 1s produced or consumed They do not allow transport of quantities They are used when the input entities concentrations do not change see Section 4 10 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 13 of 100 3 4 Processes Processes Figure 3 define th
50. eeds to be installed Java Runtime Environments JREs are available for download at http java sun com downloads Alternatively for Windows and Linux platforms you may execute the Cell Illustrator installation package bundled with a Java installer Cell Illustrator installation packages also offer an option to install a beta version of Cell Animator This section covers only the installation of the desktop product For instructions on how to install Cell Illustrator Server see the Cell Illustrator Server Installation Manual 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 6 of 100 2 3 1 Windows 23 2 Java bundled Installation 1 Uncompress CI3 0x_wj zip in a temporary directory You can use WinZip www winzip com 2 Click on CI3 0x_wj exe to install the program Java Virtual Machine JVM will also be installed 3 Launch Cell Illustrator from the Start Menu Non Java bundled Installation 1 Uncompress CI3 0x_w zip in a temporary directory 2 Click on CI3 0x_w exe to install 3 Launch Cell Illustrator from the Start Menu 2 3 3 Linux 2 3 4 Java bundled Installation To install Cell Illustrator 1 Copy CI3 0x_lj bin to a temporary directory 2 Execute the command chmod x CI10x_lj bin 3 Execute CI3 0x _lj bin file Java Virtual Machine JVM will also be installed To launch Cell Illustrator 1 Go to CI folder 2 Execute the comman
51. eeeeeeneeeneeeeeeeeneees 21 4 BIOLOGICAL PROCESS MODELING WITH CELL ILLUSTRATOR 24 fF ly COP AG ANON rice san aa cous A 24 42 FANSIOCAUON rena ETON 26 Ao MAMI SCHUM ea a A 2 AA COMDE anaa aa E de amactae tae ecanne tana 29 49 Sale HI Za lOMsaavessiietetec tater icaa tant thea eee toa cuee a 32 4 6 SlOChaSlGC etramenZaAllOM essee TAR 33 Af SUSIOM Tetramenzatli N crier testis See ol eee inden bash eo cttene nace cnet ee 34 46 DA ALON senes a acwanaiteundaaietsdannas eneeearnuisacuesebiaateane 35 A ONDION eae ero rte nm ct Se te Ot eNOS re Nr er ne ee 37 AO CAlaly SIS cenere e ERS EE 39 5 CELL ILLUSTRATOR USER INTERFACE AND MODEL CREATION 43 Oe PACES MIC PS rrasa O SZ WIOGINY a Model ctecsesncitcetraceteni asain beeen satiate netted pe cte duane ee 5 3 Change Element Properties 0 0 0 0 ccccccccceccceeeceseeeeeseeeeeseseeeseueeaeeesaee 45 5 4 Element Settings Dialog and Element Lists Dialog cceeceeeeee eens 47 5 5 Biological Elements and Pathway Fragments ccccseceseeeseeeeeeeeeees 54 56 SCNPUEdIto and PNUIS SCMPIS corseira n EE E 58 9 Preterences DIGIOG awa sicorsceriusedeiinsdee Goniearevenle eee relate te ance ts 62 5 6 Slap Way OUL Dialog rieiceactee teeter tdi E iauuieietiacihe Atelc 63 2002 2008 GNI Ltd and Human Genome Center Institute of Page 2 of 100 Medical Science The University of Tokyo All rights reserved 6 SIMULATION cioni 64 GS ImMUlalOn CONCEP ens A 6
52. een proposed as shown in Figure 1 4 We will represent this mechanism as a model with Cell Illustrator Fas ligand m1 cell membrane Fas receptor Apoptptic Stimul Apaf 1 m17 cece FADD m4 Hi Bax Bad Bim n Procaspase 8 m5 Bc 2 Bc I XI dATP ATP Bid i terminal 2 m11 m13 Cytochrome c m18 C e e e Mitochondr ion m16 Caspase 8 Procaspase 3 m25 Nei m8 A SE o a 9 P c m23 m20 2 d aspase 3 w Q t 0 a 9 m21 DFF i im30 Cleaved DFF45 m33 34 35 nuc leus aA B igomer of DFF40 P i m36 m22 DNA fragmentation 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 90 of 100 Figure 1 Proposed steps of apoptosis induced by Fas ligand Fas ligands which usually exist as trimmers bind and activate their receptors by inducing receptor trimerization Activated receptors recruit adaptor molecules such as Fas associating protein with death domain FADD which recruit procaspase 8 to the receptor complex where it undergoes autocatalytic activation Activated caspase 8 activates caspase 3 through two pathways the complex one occurs when caspase 8 cleaves Bel 2 interacting protein Bid and its COOH terminal part translocates to mitochondria where it triggers cytochrome c release The released cytochrome c binds to apoplectic protease activating factor 1 Apaf 1 together with
53. efault parameters is inserted between them Also discrete entities can not connect to continuous processes see Element Combinations in Section 3 5 Generic entities can only connect to generic processes Each connector has up to three labels above the connector 1s its name below is its firing style and below the connector is its threshold if threshold firing style is enabled or the rule if rule firing style is enabled see Figure 2 The threshold is a minimum value for the activation of the connector see Simulation Concept Section 6 1 5 2 Modify a Model The user can move any element within the model 1 e change its location To carry out this task click on the Selection Mode button from the Edit toolbar Then click on the element in the canvas and drag it to move it You can also drag the mouse to select multiple elements and then right click the mouse to bring up a popup menu which allows to you to arrange copy cut duplicate or delete the selected elements A mouse right click on a selected single element will bring up a customized popup menu with additional options to modify the element The Edit toolbar contains buttons for frequently performed editing operations This includes d Edit Parts Organize and use a repository of reusable model parts see Section 8 3 g e Create Frame Create a background frame lg Create Note Annotate the model e Manual Move Move th
54. es a problem of non firing processes This may happen when the value associated with an entity is smaller than the calculated change of this value delta within a given sampling interval If this happens the process does not fire at all This behavior is an inherent consequence of the classical Petri net formalism However it may lead to an unexpected unrealistic simulation behavior e g when the kinetics of catalyzed reactions in biological systems are modeled a catalyzed reaction may not occur at all even if a substrate and a catalyst are present The Weak Firing can be enabled disabled for continuous or discrete processes By default the this option is off which is in agreement with the Petri Net model but is sometimes difficult to understand for normal biological reactions Cell Illustrator has several commands to run a simulation If the Play Play with Animation Fast Play and Max Speed Play buttons in the Simulation 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 66 of 100 toolbar are clicked the simulation will run until the end Clicking on Step Play gt and Step Play with Animation buttons executes a simulation step corresponding to one sampling interval If a simulation process is started with the Play or a Play with Animation button the program attempts to execute the simulation at the speed 1 pt per 1 second regardless of t
55. f Tokyo All rights reserved Page 48 of 100 element setings Entity Figure Biological Name let Variable ni Value Type Double Physical Properts unknown o Unit fanit Initial Value bo y valuate Script Once fon Maximum value Infinite Minimum value b Public off Save Log off Show Element fon Show Mame fon Show Variable jon Show Value jon References Select Add Edit Reference comments Figure 54 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved P age 49 of 100 5 4 2 Processes In the Process List in the Element Lists Figure 55 you can change the name of the process to reflect its biological meaning Also you can set the Kinetic Script and Delay Script property The Kinetic Script and Delay Script fields define respectively the speed add update function and delay parameter of the processes The Kinetic Value and Delay Value columns display current simulation values xl Firing Style Kinetic Script Kinetic Value Delay Script Delay value Log isi a T i CE 8 kz a E Vv Show Options Figure 55 If a process 1s selected in a canvas the Element Settings display the properties of the selected process on two sheets the Process Sheet and the Figure Sheet Figure 56 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All
56. for the simulated model The SECG allows the change and extension of the common module see the EngineBase options in Simulation Setup SECG menu This enables the usage additional functions scripts or performing operations which would normally would be impossible in the Cell Illustrator for example saving and using previous values of the entities 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved P age 74 of 100 The source code of the engine generated by SECG can be exported to file in Java or other languages The source code file can then be edited which allows for better control over the simulation Also the simulation source code can be controlled and executed in other program which is useful for performing parameter optimization or generating some advanced statistics from the engine run The exact reference for advanced usage which contains description how the SECG framework works and detailed information of the content of each source code file is obtainable from Masao Nagasaki masao ims u tokyo ac jp or check http www csmlL org Note Profiting from the SECG engine possibilities to the full extent requires some software development skills 6 5 Remote Simulation Servers In Cell Illustrator simulations are normally executed at your desktop machine using the default simulation engine which is tightly integrated with the remaining program components These
57. g For example you must specify the Coefficient1 Coefficient2 Standard Deviation and 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 51 of 100 Connector Stoichiometry for the Stochastic Mass style The most general style is Custom If a process has the Custom kinetic style you need to specify directly its kinetic script i e the formula that governs the value change For other kinetic styles the kinetic script is predefined and you can only set the values of parameters in the equation The Delay option is enabled only for a discrete process see Section 3 4 and it takes a positive real value or zero for continuous processes the delay is always set to 0 A discrete process will fire just after its delay time if the process is still active Otherwise it can lose its chance to fire The delay is specified in Petri net time pt By default the Activity property of a process 1s set to true However you can write a script in the Activity field to turn the process on and off in the course of a simulation For an example on how to write a script see Section 5 4 4 In the Figure Sheet you can modify the display properties associated with the process 5 4 3 Connectors x ee Se a a CT a cA Z oo l o M a iboe OOOO OOOO esd o o Show Options Figure 57 In the Connector List of the Element Lists dialog the Name
58. gand concentration lt 1200 1000 z z 800 an Autocatalytic di me a z 3 600 o cae i absent 5E 400 HF D 200 O O O 200 400 600 800 Concentration of Fas ligand 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 93 of 100 Figure 3 Simulated relationship between the amount of DNA fragmentation and the Fas ligand concentration At higher concentration of Fas ligand the direct pathway from caspase 8 to caspase 3 contributes to the fragmentation To examine the effect of autocatalytic process of caspases DNA fragmentation is simulated for both cases of the presence and absence in this process There are two pathways from activated caspase 8 to caspase 3 one through several steps including the cytochrome c release from mitochondria when the concentration of activated caspase 8 is low and the direct one to caspase 3 when the concentration of activated caspase 8 is high 3 We assume arbitrarily that the direct pathway starts when the concentration of activated caspase 8 1s larger than 30 Reportedly the removal of the Bid by gene knockout method increases the resistance of liver cell apoptosis by Fas ligand while it does not affect the apoptosis of thymus and embryonic cells If the second pathway is included in the scheme DNA fragmentation increases slightly especially when the Fas ligand concentration is high Figure 3 However
59. gical Properties Tab The Biological Properties tab available in the Entity Process Settings dialog box enables you to set additional biological information for the selected entity or process such as location GO ID list of Pubmed ID s For a process you any select the Process type from a list of predefined choices Please note that changing the process type will assign the graphical symbol image of this process to the figure on the canvas For an entity a list of references to external databases can be defined e g UniProtID EntrezGeneID MIMID etc 5 4 5 External References Dialog Box Using this dialog box you may define for each model element a list of references to external databases or vocabularies The defined links can then be opened and viewed in a web browser 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 53 of 100 5 5 Biological Elements and Pathway Fragments While you can use the menu options and toolbar buttons described in Section 5 1 to add elements biological pathways can be created faster and more intuitively by dragging and dropping appropriate elements from the dialog boxes that support building of pathways These dialog boxes give the user the access to many predefined elements or pathway fragments and the ability to insert these fragments to the active canvas 5 5 1 Biological Elements Dialog The Biological Elements dialog conta
60. gure 39 x Copy To Clipboard Figure 39 If the speed of the separate process strictly depends on the value of p53_mdm2 say m1 this can be expressed by a respective equation defining the speed parameter of the process In Figure 40 m3 20 for the model where all elements are of the continuous type 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 36 of 100 E sa_mdam 3 p mi c1 pi eg _mdm Mu s00 maz20 00 50 0 mdm2 mg 0 0 Figure 40 The results of the simulation are presented below Figure 41 x Copy To Clipboard Name coor m pamm B M Figure 41 For models based on discrete elements it is difficult to create a model that fulfills this condition 4 9 Inhibition A specific drug sometimes inhibits activities of transcription The transcription model in Section 4 3 is modified with an inhibitor activity as follows Step1 Create an entity and set its name e g doxorubicine Step2 Create a process and connect it to the doxorubicine entity with the process connector Step3 Connect the doxorubicine entity to the transcription process with the inhibitory connector Step4 Do not change the threshold of the inhibitory connector the default value is 0 The type of the doxorubicine entity can be either discrete or continuous With the model in Figure 42 one can simulate the situation wher
61. he Simulation Speed option in the Simulation Settings dialog On the other hand for simulation started with Fast Play and Max Speed Play Cell Illustrator attempts to execute simulation at the rate defined by the Simulation Speed For example if the Simulation Speed is set to 100 then the simulation will proceed at 100 Petri net time units per 1 second of real time provided that your hardware is sufficiently fast You may need to increase the Sampling Interval at the cost of result accuracy if the simulation is too slow The difference between Fast Play and Max Speed Play is that the canvas is not updated until the simulation end if the latter option is executed Thus you should use the Max Speed Play for time consuming models The Log Update Interval option defines how often entity values are saved in the simulation log To track value changes for a specific entity you need to set its Log option to the on value The Chart Update Interval defines how often simulation charts are updated This option has no meaning for simulations started with the Max Speed Play command fc Simulation Settings Sampling Interval pt ba oo Chart Update Interval pt Booo Log Update Interwal pt ooo Simulation Time pt hooo o Continuous Weak Firing of al Discrete Weak Firing of al Firing Accuracy fi DE 10 Simulation Speed ptis ao a0 fo 100 Restore Defaults Close Figure 74 6 2 2 Simulation Log The results of each simu
62. hen the operation completes Since CIL stores the model in the same format as in the original project file you can open a log file with Cell Illustrator and edit the model However saving the edited model into the original log file will erase the time series and any other simulation history information Therefore you should use the Save As command of Cell Illustrator if you would like to recover the model from a log file for further editing 6 2 3 Simulation Charts You can visualize the simulation results with concentration or amount versus time graphs Charts can be defined in the Chart Settings dialog Figure 77 Click on the button in the Property Dialog toolbar to bring up the Chart Settings A table at the left side of the dialog displays all entities in the active canvas The Chart Tree at the right side displays all defined charts for the model in the active canvas Select an entity in the table and click on the Add Selected Entities button to create a chart that tracks the entity value over time A newly created chart is added to the tree To add an entity to an existing chart highlight the chart name in the tree select the entity in the table and press the Add Selected Entities button The maximum number of entities in one plot is 8 You can also delete charts Highlight names of charts to be deleted in the Chart Tree and press the Delete Selected Charts button Plot Settines Add Selected Entities Remove Selected Plots e
63. ine Server and open selected models in CI workspace 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 56 of 100 x Project File Sort Search Previous Mext CI File Server My Projects Bh oly i project Refresh Close Figure 62 To open Project Manager Dialog choose icon from the top toolbar 5 5 3 CSMLDB Search Dialog The CSMLDB Search Dialog Box is an interface that allows for submitting queries to the remote CSMLDB database To access the CSMLDB database the user must be logged in to the CI Online server Using this dialog you can search in the CSMLDB database for molecules entities and or reactions processes and import them into active canvas To open CSMLDB Search Dialog select icon from the top toolbar 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 57 of 100 Search CSMLDB Academic x Entity Process Search For p53 Find Search In Type FN v0 i J Order By T l EE Displaying results 1 50 of 457 000583 o Gooio7s pS 3th mR MA Goososs TP53BP1 H 00376 TPS Gooso fresat Goois PPsafpo TPS3 2 GO0101S a rat TRA e0i5is0_ aoa 5 TPS SBP 1 m S MRMA a be postr 5 1 MmR MA 457 Molecules For pS a ee J With merge Create new canvas Figure 63 5 6 Script Edito
64. ings dialog 6 3 Viewing and Analyzing Simulation Results 6 3 1 Simulation Charts Entity values in charts are color coded Right click on a chart to open a pop up menu with options to customize the chart Figure 78 x Rasterize Copy To Clipboard 100 77 e0 ao fon 60 50 4 SAVE aS 1 e 4c a ee ae Prine w a i E zoom In b su l i oom uk 104 o gt ee eee d SP ee eee ee os Suto Range 0 250 500 T50 1000 1250 1500 1750 2 000 Time Figure 78 e Properties opens a dialog box in which you can set the chart visualization details e Save as Turns on and off the tool tip for the graph If the option is enabled you can see the value of the data point once you move the mouse cursor over it e Zoom In Zoom Out Turns on and off the zoom in feature You can drag the mouse cursor over an area in the chart to zoom into it e Auto Range adjusts the range of x and y axis 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 71 of 100 e Copy to Clipboard button Copies the chart to the system clipboard You can export the chart to another application e g Microsoft Word or Excel in this way e Entity List enables you to change the visualization and color of each single entity plot 6 3 2 Simulation History Dialog For your model you can run simulation many times changing the simula
65. ins the Entity Sheet the Process Sheet and the Others Sheet The Entity Sheet contains biological elements that will contain a certain quantity e g mRNA and protein Figure 59 c Biological Elements Element Type Element SQ A P Continuous C Discrete Generic Scale fio 100 E SBI Figure 59 The Process Sheet contains biological elements that denote biological processes e g phosphorylation and translocation see Figure 60 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 54 of 100 d Biological Elements x Figure 60 The others sheet contains useful pictures of cellular components e g cell and mitochondrion Figure 61 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 55 of 100 Biological Elements x Figure 61 5 5 2 Project Manager Dialog CI Project Manager is a user interface to pathway models stored on CI Online Server To access the remote CI Online Server the user must be logged in In this dialog box you can Organize your models and projects stored on Cell Illustrator Online Server Share your projects with other users and use projects shared by others See the CI Reference Manual for details Browse and search through pathway libraries defined on the CI Onl
66. l not be created during the simulation A CIL file is created if at least one entity or process is selected for logging You can set the Log flag for all entities or processes at the same time by setting the Column selection box in the Options panel and clicking on the Log column The Row selection checkbox in the Options panel must be unchecked At the simulation time the data is logged time points defined by the Log Update Interval setting in the Simulation Settings The log interval value needs to be selected carefully so that an appropriate number of steps is logged during the simulation In general the number of steps that will be logged is given by the following formula SIMULATION _ TIME N _ STEPS 7 LOG _ INTERVAL It is recommended to select the log interval in such a way that the number of steps will be less than 100 000 steps Logging hundreds of thousands of steps will cause the CIL file to be very large 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 68 of 100 The formula for the number of steps V_ STEPS is accurate if entities only are logged In case processes are logged as well each simulation step in which a process state changes 1s added to the log This means that the log may include many additional time points besides the time points defined by the log interval In such a case the number of logged steps can be greater than the value given
67. l rm Figure 77 6 2 4 Run Simulation Once the setup described in the previous section is completed you can start and control the simulation using the VCR like controls in the Simulation toolbar The toolbar allows you to start and pause the simulation Also you can step through the simulation with the Step Play 1 gt button The buttons and invoke simulation with animated tokens moving between entities The simulation can be stopped at any time with the Stop and Initialize M button These simulation commands are also available in the Simulation menu in the Menu Bar 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 70 of 100 During the simulation active components in the model are highlighted in red You can also see the values of the entities updated during the run Charts defined in the Chart Settings dialog are displayed and updated with the frequency specified by the Plot Update Interval option in the Simulation Settings dialog The simulation results for the logged Entities can be saved in a CSV Comma Separated Values file which can be opened outside Cell Illustrator in a text editor or a spreadsheet After your simulation is completed press the Save Simulation Log button You will be prompted to enter the name for a CSV file that will store the log The logging frequency is defined by the Log Update Interval option in the Simulation Sett
68. lation are presented below Figure 27 x Copy To Clipboard wane e mRNA_pS3 m2 Ma mRNA_pS3 m1 T7 Ml mRNA_pS3 m3 _ El Figure 27 4 4 Complex In a cell a set of proteins binds and creates a complex A model for this can be created with the following steps Step1 Create three entities and set their names e g p53 mdm2 p53_mdm2 and set their values e g 100 50 0 Step2 Create one process and changes its name e g complex Step3 Connect p53 entity and mdm2 entity to the complex process Connect the combine process to the p53 mdm2 2 entity 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 29 of 100 Step4 Set the combine speed by editing Speed Add options of the complex process the default value is 1 0 In Step1 and Step2 9 combinations of element types in the pattern p53 mdm2 p53 mdm2 complex process are allowable 1 e discrete or continuous discrete or continuous discrete or continuous discrete and continuous continuous continuous continuous Figure 28 p53 mi pS3_medm2 m pl ca 100 meme 50 psa m4 43 pS3_medm2 m pa c 100 mem 50 0 p53 mi 3 3 pha _mdm ma p3 co 100 0 mem me 50 paa m10 nid 100 0 mdm2 _ mA 50 0 c15 PS p53 mia p5 100 0 mem mje 50 0 c12 pS3_medm2 mi2 aomdm2 m15
69. lation can be logged and saved into CI log files CILs The CIL files can be viewed and analyzed in the Cell Illustrator Player program The simulation log can also be exported to a CSV file and then analyzed with an external spreadsheet application 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 67 of 100 A CI log is an XML file that consists of two parts the input model and a time series of values for the logged properties The time series data may include entity values and or process state firing waiting to fire due to the delay property not firing You can define the entities and processes to be logged in the Element Lists dialog by selecting the Log option as it is illustrated below xi entity envi Dabe emvrome 0 e e r re e A e a e e E E B a a entity Lm Daubie imtme _lenvironme O r e e entity Lm Daubie mkom kneticpar r e e entity Im_ Bouble imkem kneticpar O a r e e LAATTI KAITI erity im Dank recor O N e entity Im_ BouBle imke kinetic par T C O a a city Bale inten Be CIN nD ery Pele ine eraome TP entity aii sae Hm Ta m kinetic par Show e cee a 4 4 PAA i i 4 4 ite aa 4 4 4 4 Close Figure 75 By default the Log option of all entities and processes is off This means that nothing will be logged during simulation and a CIL file wil
70. lication Only one CI Player application window can be opened at a time If CI Player is already running at the time you press the Replay button the selected files will be loaded into the CI Player application window 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 72 of 100 6 3 3 Exporting Simulation Logs Once a CI log file CIL 1s created the simulation results can be exported to a CSV Comma Separated Values file using the File Export Simulation Log command The CSV files can be viewed analyzed and edited by a third party application e g a spreadsheet or a text editor An exported CSV file will contain time series that remains in the in memory log buffer of Cell Illustrator at the time the export command 1s executed In case of a large log the buffer may contain only a part of the log the time series data is being saved to a hard disk while the simulation is running You can set the size of the in memory log buffer in the Preferences dialog in Cell Illustrator To export the whole log load it into the CI Player and execute the Export Simulation Log command there 6 4 Running SECG Simulation SECG Simulation Engine Code Generator is a new mode of running simulations in Cell Illustrator SECG is based on the same HFPNe model as the standard simulation engine described above however it is faster more customizable and adds some viewing and running
71. llular structures like mitochondria cell nuclei cells or biological phenomena like transcription and translation An entity Figure 1 contains a value which is interpreted as its quantity or concentration You can input a more meaningful name and value for the entity Also each entity has an associated variable e1 e2 e3 representing its quantity 30 123 2 ATGC respectively The mathematical equations used for simulation are expressed in terms of these variables 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 12 of 100 There are three types of entities discrete entity continuous entity and generic entity I name variable cAMP el enzymeA z mRNA e3 I i l quantity l 30 1253 2 AUGC l i Discrete Continuous Generic i Entity Entity Entity Figure 1 Discrete entity holds an integer quantity Integer or Long In contrast a continuous entity has a continuous e g real fractional number as its quantity Double This type of entity is used to represent concentrations e g the number of ions or enzymes Other type of entity is a generic entity It holds string or logical value String or Boolean 3 3 Connectors Connectors are used to connect entities to processes input connectors or processes to entities output connectors There are three types of connectors Figure 2 process connectors inhibitory connectors and association conn
72. me 4 the value of el is the same as the threshold of cl 1 e 2 At that time the process p1 cannot be enabled and el and e2 reach a stable state The results of the simulation are presented below Figure 9 x Copy To Clipboard Figure 9 The same conversion process can be applied to the discrete model in Section 3 7 Figure 10 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 20 of 100 me 1a i T 18 Po Figure 10 The equations for quantity changes are _ dmi _9 dm2 dt dt The results of the simulation are presented below Figure 11 x Copy To Clipboard Figure 11 3 9 Continuous Elements Example 2 In the previous two examples discrete examples were converted into similar models with continuous elements For kinetic based modeling one should use a continuous model rather than a discrete one In the following model Figure 12 the consumption ratio of el 1s proportional to the value of el 1 e the speed of process p1 follows the mass action kinetics el mi e2 m2 i ci p1 c2 Or QO 10 0 mi o 05 0 0 Figure 12 2002 2008 GNI Ltd and Human Genome Center Institute of Page 21 of 100 Medical Science The University of Tokyo All rights reserved The difference between the present model and the model in Section 3 8 is that the speed of p1 is changed from 1 to 0 05 m
73. mended to run a simulation after a successful model import to verify whether the results of the converted model are identical to the simulation results of the original CellML model Before running the simulation the user should check and set the proper Simulation Time and Sampling Interval in the Simulation Settings dialog box 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 85 of 100 8 6 Import of CSO and BioPAX models Cell System Ontology 3 0 CSO3 0 www csml org online services csml 30 ontology is an OWL format for storing biopathways information It is based on CSML3 0 which makes it a very complex and powerful format BioPAX level 2 www biopax org is most popular ontology for storing biological pathways data BioPAX Level 2 covers metabolic pathways molecular interactions and protein post translational modifications In Cell Illustrator you can import models from CSO and BioPAX files into CI workspace During the import operations the models are converted to CI models and can be edited analyzed simulated saved as any other models Each model opened in the workspace can be exported to CSO ontology 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 86 of 100 9 Cell Illustrator Player Once a CI log file is created during a simulation it can be opened in CI Player
74. ments Example 1 In the previous two examples discrete entities and discrete processes have been used However in many cases entity values should be treated as continuous variables rather than discrete variables The model in Figure 8 has been created by conversion of the discrete model in Section 3 6 Figure 8 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 19 of 100 More specifically the differences between the models in Figure 4 and in Figure 8 are as follows Step 1 The type of entities el and e2 1s changed from discrete to continuous Step 2 The type of process p1 is changed from discrete to continuous All other parameters are the same as in the discrete model in Section 3 6 Note that the Delay parameter of p1 has no meaning now since it is not applicable to a continuous process For the continuous processes the delay parameter is always zero As described in Section 3 6 the threshold parameter of the input connector works like the threshold of the activity of process p1 In this example at time 0 the value of el is 10 and the threshold of cl is 2 and the process is enabled When all input connectors are enabled the connected continuous process is executed with no delay Since pl is connected to el with the quantity m1 and its output is connected to e2 with the quantity m2 the equations for quantities change are dmi_ dm2_ dt dt At ti
75. ments dialog see Section 5 4 4 Each entity has three labels associated with it for you to edit by double clicking see Figure 1 On the upper left is the Name of the entity for associating biological concepts with the entity The upper right label is the Variable representing this entity It is used in mathematical equations to define the rate of reaction in the model The lower left label represents the current value of the entity You can edit it to enter the initial value for the selected entity Note that for discrete entities this value must be an integer 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 43 of 100 A process can have up to four associated labels see Figure 3 At the upper left is the Name of the process The lower right label is the Kinetic Script of the process The upper right label is the Firing Style of the process and the lower left label is the Delay of the process _ To adda connector to link two elements select a connector type from l and icons in the Edit toolbar above the canvas pane and then click on the element where the connection originates Next click on the element you want to connect to The connector will be automatically drawn to connect the two elements Note that a connector links entities to processes not entities to entities or processes to processes If you draw a connector between two entities a process with d
76. modes The basic concept in SECG consists in generating a computer program which execution will perform the simulation Also the SECG module allows exporting the source code of this program this allows running the simulation from other program and enables controlling the simulation behavior and data in a more precise way The simulation can be run using SECG module by selecting Simulation Run SECG Simulation in the menu The simulation is performed outside the Cell Illustrator program and many simulations can be done simultaneously The drawback of SECG simulations is that they don t provide any control of the execution of the simulation and may consume lot of computer resources but at the same time those simulation are executed in much shorter time than the simulations which are run in the Cell Illustrator workspace SECG can output the results of the simulation as charts and or log file this options are customizable from Simulation Setup SECG Moreover SECG enables inserting reference data into the simulation plot which enables visual comparison of simulation results versus reference data contained in CSV file this simulation mode can be run by invoking Simulation Run Reference Simulation Note The javac compiler that is included in the JDK 1s required to run SECG simulations The JDK Java Development Kit must be installed on your computer JRE is not enough to run SECG 6 4 1 Passing Parameter simulation The SECG module
77. ne account 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 5 of 100 Run Cell Illustrator https cionline hge jp cihilesernver 7 Password lowes Remember Password cimaxmummenone o T Inthe CI Startup dialog enter the login data that you obtained during the registration and press Sign In Note It is recommended to select the Remember Password option After a successful login the login and password will be stored on your local computer and you will not need to enter the login data again After this CI Online is downloaded from the server This may take several minutes CI Online is a client server solution that enables the user to create manipulate and simulate bio pathway models locally within CI workspace and to store and organize them on the remote CI Online Server Additionally CI Online gives the user convenient access to libraries and databases of bio pathway models In CI Online Cell Illustrator is the front end where the user prepares a pathway and simulates it The Cell Illustrator client communicates with the CI Online Server over the internet HTTP or HTTPS protocol The CI Online Server is a place where the user logs in to store and manage his pathways and to share them with other users 2 3 Install and Run Cell Illustrator desktop version In order to run Cell Illustrator correctly Java version j2sdk 1 5 0 or higher n
78. ning simulation press the Stop button After this command the program will attempt to stop the simulation at the server and download the output files of the interrupted job The Delete button removes the simulation from the list without downloading the output files 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 76 of 100 Remote Simulations n Submitted Simulations notch gon igrafitonDebug Simulation Error Nothing is selected For logging in the in notch gon sigrafitonDebug Finished remote simulation hotch gon ___lgrafitor natch gan grafiton Finished remote simulation running 14 Estimated time left For execution 317 s di Remote Simulations Details notch gon x Remote Engine URL Input File Output File Start Time Job Status Job Message Job Progress grafitoa hittp arafiton faspl com pl testAxis services CIRemoteSimul Funning E 14 Close Figure 81 Submitted Simulations Stop Delete fe Details Close After a remote simulation job is completed its log file is downloaded from the server and saved in the output directory of the model The simulation log appears in the list in the Simulation History Dialog You can open view and analyze the simulation log in Cell Illustrator Player the same as you can with any other log file When the download of the log file is completed the message
79. nstitute of Medical Science The University of Tokyo All rights reserved Page 42 of 100 5 Cell Illustrator User Interface and Model Creation At Cell Illustrator startup a single main window pops up You can create open save and print a biological pathway model file using options in the File menu Figure 52 This window has the Menu Bar at the top that allows you to manipulate the model diagram Cell Illustrator Online 4 0 demo localhost Oj x File Edit Element Simulation View Analyze Dialog Window Help BOs ck OCOD 1 1 total Free 1016MB 880MB Figure 52 To create a new model the first step is to open an empty canvas window for the model which can be accomplished by either 1 clicking on the top icon in the left toolbar Create New Canvas Window 2 clicking on the File in the Menu Bar and selecting the New option 5 1 Add Elements To add a new element an entity or a process select one of p Cy E and icons from the top toolbar and then click on the canvas workspace where you want to place it You may continue and place more components of the selected type on the canvas Alternatively you can press the right mouse button anywhere on the canvas and choose the desired element from the popup menu using Insert Entity Insert Process function The new element will be placed at the center of the active canvas Lastly you can select a predefined element in the Biological Ele
80. nstitute of Medical Science The University of Tokyo All rights reserved Page 9 of 100 To run a simulation on a sample gon file e g apoptosis gon Open your command line terminal Change to the CI installation directory The example script should be run from the installation directory Type the command ClEngine inputFilePath e g ClEngine sample apoptosis gon If you are experienced in Java please feel free to modify the example script and customize it to your needs 2 6 CI Memory Customization 2 6 1 Desktop Version Using the auxiliary application CI Memory Configuration you can set the maximum memory for both Cell Illustrator applications CI and CI Player The application can be found in the installation directory 2 6 2 CI Online When starting CI Online you can set the maximum memory for Cell Illustrator in the Login dialog box 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 10 of 100 3 Cell Illustrator elements CI is designed to model biological pathways using three types of abstractions entities processes and connectors 3 1 Concept Overview The modeling and simulation engine of Cell Illustrator is based on an extension of the Petri net methodology A Petri net models a system of conditions and instantaneous events their dependencies and mutual synchronization The Hybrid Petri net with extensions HFPNe employed in th
81. nt csv Ci x FA pr Connector List 183 z _ Print 4rea BEB Simulation Log a33 F Ea File Edit View Insert Format Tools Data Q Print OOO aava Window Help R x J ext T bed f Arial 10 BZU at z dy m20 m21 10000 R2 m22 l repu iga EE i Variable Initial Valu Eve gt gt ae Fas ligand Double m3 0 r CS eevafi ft 3 pro caspasDouble m5 39 039 a NOE Pi oma 4 DISC Double m6 60 961 oe l p complex Double mi 0 5 traidd 6 caspase8 Double m8 0 BID Double m11 100 8 BID C term Double m13 0 saa 9 cytochrom Double m16 0 oot 20000 10 Apaf 1 Double m17 39 039 11 dATP Double m18 39 039 complex Double m20 0 13 pro caspasDouble m21 100 complex Double m22 0 m23 DAI KETA E e ev e 0 Export entity list done total Free 98MB 26ME Figure 85 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 81 of 100 8 Making Complex Models This chapter presents features of CI that are especially useful for large scale modeling 8 1 Grouping and Ungrouping When editing a model it is often convenient to treat a group of elements in the canvas as one element This can be accomplished with the Group command This command is accessible in three ways via the popup menu in the canvas Element Group in the main menu and the Group icon in the Edit toolbar Once elements are grouped you can select and move them around easily ju
82. o the Mass kinetic style can be found in the section 3 4 Processes The results of the simulation are presented below Figure 33 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 32 of 100 x Copy To Clipboard 10 0 lt Mame Color aaa po3 tetramer 75 ia hd pS3imoanamer W o Figure 33 4 6 Stochastic Tetramerization Normal biological processes e g dimerization or tetramerization are not deterministic but stochastic To improve modeling of the biological processes one can use the Stochastic Mass kinetic style of the process Starting from the previous model the stochastic tetramerization process can be modeled as follows Step 6 Change the Kinetic Style of the process to Stochastic Mass Set the Standard Deviation to 0 2 This parameter controls the distribution of the randomly generated process speed Step 7 Make sure that other parameters are the same as in the previous Tetramerization example 1 e cl Stoichiometry is 4 pasimoanomer _ mi pi cI posttetramer _ me o 0 SMass m1 0 0010 0 2 r Figure 34 In the model shown in Figure 34 the speed of the tetramerization process represents stochastic behavior Parameters related to Stochastic Mass kinetic style are described in section 3 4 Processes The example results of the simulation are presented below Figure 35 Note that subsequent simula
83. ociated with a process 1s its Kinetic Style If the process kinetic style is Custom then the value change after the process fires 1s the same for all connected input output entities For all other kinetic styles the value change at each entity depends on connector specific parameters such as Connector Stoichiometry and Connector Speed Thus it may be different for different input output entities 3 5 Element Combinations All combinations of discrete and continuous CI elements are presented in Table 2 and Table 3 Table 2 Output Connectors L v t L v t Transport quantity Transport quantity Sequentially v t through the POVANOWEQ v t through the transport quantity Connector at d t Connector at each v t through the intervals d t interval Connector 56660006 Not Allowed The variable m t represents quantity or concentration of the entity at time t The variable w t is the threshold of the corresponding connector The constant d t is the delay time of the corresponding process and the function v t is the speed of the corresponding process at the time t Note that CI checks and prohibits the connections that are not allowed during model creation 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 15 of 100 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 16 of 100
84. og Window Help E k OO L AVRR SS BES E F H 2 apoptosis_20 gon Fasigand mi ozs avg Hl PANH AD 75 noe Fas ligand tiene m2 0 OA Gt and jand 000001 4 m2200 0 Display Navigator done total Free 67MB 16MB Figure 82 2002 2008 GNI Ltd and Human Genome Center Institute of Page 78 of 100 Medical Science The University of Tokyo All rights reserved 7 2 Finder With the Finder dialog you can search for specified elements in the active canvas Figure 83 As with the Navigator the Finder becomes very useful when the model in the canvas is large or very complicated x Search For iF M Regular Expression Case Sensitive Search In J Search Names M Search Comments J Search Variable Names Search Variable Names in Script Found 4 entity Fas ligand trimer Fas receptor complex entity Fas ligand entity FADD entity Fas ligand trimer Close Figure 83 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 79 of 100 7 3 Print and Print Area In CI you can output the whole canvas a part of active canvas or the simulation charts to a printer or an image file in the Postscript PNG or JPEG format Figure 84 This functionality is available via Print and Print Area options in the File menu laxi Fie Edit Element Simulation view Dialog Window Help Print saan a mo 4 ne Wy H 9 amp ow vA
85. oi g Figure 18 The results of the simulation are presented below Figure 19 d Chart 2 my x Rasterize Copy To Clipboard Figure 19 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 25 of 100 4 2 Translocation In a cell mRNA and proteins move from one compartment to others e g from nuclei to cytoplasm and vice versa A model for this can be created with the following steps Step1 Create two entities and change their names e g p53_ nuclei and p53 cytoplasm and set their values e g 100 and 0 Step2 Create a process and set its name e g translocation Connect the p53_ nuclei to the process and connect the process to p53_ cytoplasm Step3 Set the translocation speed by editing the Speed Add option of the translocation process the default speed is 1 0 In Step1 and Step2 possible combinations of p53_ nuclei translocation p53 cytoplasm triple are discrete discrete discrete discrete discrete continuous continuous discrete continuous continuous discrete discrete and continuous continuous continuous Figure 20 poa_nuclei _ mi ef translocation eg PSs_eytoplasm _ m2 100 aM 1 10 pas_nuclel m3 c5 translocation cg Pa3_eytoplasm _ m4 100 ig 110 0 0 po3_nuclel _ m5 c7 translocation eg Pas_cytoplasm _ m6 100 0 a 1 1 0 0 0 p53 nuclei _ m7 eg translocation ag Pas_cytoplasm ma 100 0 ae
86. ovided Note that the Dialog menu contains all options provided in this toolbar The actions for each button are as follows Toggle Element Lists Visible Status open or close the Element Lists dialog see Section 5 4 g z z Toggle Element Settings Visible Status open or close the Element Settings dialog see Section 5 4 TE Toggle External References Visible Status open or close the External References dialog see Section 5 4 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 45 of 100 rs Toggle Navigator Visible Status open or close a small view of the entire model to navigate the model in the canvas window see Section 7 1 X Toggle Preferences Visible Status open or close the Preferences dialog to change the graphical display properties of the various elements and global settings in this application see Section 5 7 kdi Toggle View Settings Visible Status open or close the View Settings dialog The dialog changes the view settings of the active canvas e g shows and hides entity names o H Toggle Chart Settings Visible Status open or close the Chart Settings dialog to create time series plots see Section 6 2 2 r Toggle Simulation Settings Visible Status open or close the Simulation Settings dialog to change simulation parameters see Section 6 2 E
87. pend on its Firing Style parameter and are described in detail along with other process parameters in Section 54 2 The process type determines its calculation style The continuous process has the speed calculation style the process calculates the speed of production or consumption of the entities On the other hand the discrete process has the add style the process calculates the concrete values that should be added subtracted to the entities Finally the calculation style of the generic process is update 1 e the process calculates a concrete value that will replace the entity value The Kinetic Script parameter specifies how to calculate the value change of connected entities in each of the three cases See the Reference Manual for details Each discrete and generic process has an additional Delay parameter that can be a constant or a function The firing of a discrete process occurs abruptly with the delay associated with the process The same applies to generic processes The activity of continuous processes is different The firing of a continuous process occurs continuously Consequently a continuous process does not hold the Delay parameter 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 14 of 100 To employ generic processes one must use the Pnuts scripting language Use of Pnuts is briefly described in Section 5 6 Another important parameter ass
88. ption process In Step1 and Step2 possible combinations of the pair transcription mRNA p53 are discrete discrete discrete continuous and continuous continuous Figure 24 Cell Illustrator facilitates the modeling of the path DNA gt mRNA gt Protein not only as a concentration function of DNA mRNA and Protein Global transcription and translation mechanism can be established with a string input DNA sequence intermediate product proper mRNA and product Protein This mechanism can produce simultaneously many proteins from different DNA sequences transcription cq MRMA_pS3 _ m1 transcription cq MRNA_ps3 _ m3 1 10 0 1 0 a transcription co MRNA_pos _ mz 1 10 D0 Figure 24 The simulation results of the discrete model the are presented below Figure 25 x Copy To Clipboard so mme cor aa p53_cytoplasm v 30 20 10 4 Figure 25 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 28 of 100 Normally mRNA synthesis is associated with its degradation The models presented below Figure 26 combine the transcription process with the degradation process in Section 4 1 The same speed m 10 is applied transcription eq MRNA_pssimt om transcription cARNA_pS3im3 m3 1 1 0 oh 1 0 0 0 O10 010 p4 p miso maid transcription c3 mRNA _pSsim2 me maig Figure 26 The results of the simu
89. r and Pnuts Scripts An advanced user of Cell Illustrator may need to write a custom function to specify changes of a property value With the Script Editor you can write a Pnuts script to define Initial Value property of an entity Activity Speed Add Update and Delay of a process Threshold Updater and Firing Rule of a connector 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 58 of 100 Pnuts is a script language for Java environment The Pnuts Language Reference can be downloaded from the project website at https onuts dev java net Script Editor File ifirandi lt 0 p ireturn truer else return false Figure 64 Figure 64 shows the Script Editor At the top left in the dialog is a set of function buttons At the top right is a table containing all entities in the model In the middle a script pad is placed where you can edit a script either by clicking on the function buttons and selecting entities from the table or by typing in text Remember to press the Return key at the end of the script You can check whether the syntax of your script is correct by pressing the Check button Error messages will appear in the bottom panel 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 59 of 100 5 6 1 Pnuts Script Examples 5 6 1 1 Example 1 Stochastic
90. rag and drop biological elements from the Biological Elements dialog Section 5 4 4 and link them with connectors observing the connection rules listed in Section 3 5 All element parameters are initialized with default values the initial value of an entity is 0 the threshold parameter of an input connector 1s 0 and the speed parameter of a process 1s 1 by default For any biological process to take place during simulation the process representing it needs to be enabled A process is enabled when all the following three conditions are satisfied 1 all of its incoming process and association connectors are activated 11 no input inhibitory connector 1s activated 111 the activity parameter of the process is true The default value of the activity parameter is true but one can write a script that modifies this parameter Once a process is enabled the biological process can occur Let us take a look at the simulation of a simple case A protein precursor is phosphorylated to be an active protein proteinfinactive auto phosphorylation protein active m c1 c2 10 0 2 0 mi 0 1 m2 1 0 12 Figure 71 The initial value of the protein inactive is 10 While the concentration of protein inactive 1s greater than the threshold of its outgoing connector cl which is 2 0 the connector is in the enabled state An enabled connector in this case executes the activation of the Phosphorylation The protein inactive will be converted to th
91. re requires the skills of an IT system administrator and is described in the Cell Illustrator Server Installation Manual Once the CI Server is installed you need to configure your CI Workspace to access it 6 5 1 Registering a CI Server To submit simulations to a server you must add it to the list of servers in your CI Workspace For this you will need to specify the URL address of the respective web service Optionally if the web service is password protected you will also need to specify the authorization data user name and password The menu command Simulation Setup Servers opens the Setup Servers dialog in which you can view and manage the list of servers With the Test button you can confirm that the selected C7 Server is accessible 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 75 of 100 Ix Setup Servers 7 Add engine 1 Remove Edit Name lengine URL D axis services CIRemoateSimulation User Cluser Password pene Test Hi OK Cancel Figure 80 Server Setup 6 5 2 Running a Simulation Job at a CI Server To submit the model in the active canvas for simulation at a remote server select the server name in the list available at the Simulation Send to Server menu option The simulation jobs starts and you can view its progress in the Submitted Simulations dialog that opens up To interrupt a run
92. ree examples below show how various kinetic styles processes can be used for the modeling of tetramerization The speed of tetramerization 1s linearly dependent on the mass of monomers in the first example In the second example stochastic perturbation is introduced Finally in the most advanced example tetramization is defined by a script A simple model of tetramerization can be created using the following steps Step 1 Create two entities and set their names e g p53 monomer and p53 tetramer Set the initial value of the monomer to 10 Step 2 Create a process and set its name e g tetramerization Step 3 Connect the p53 monomer entity to the process with a process connector Step 4 Connect the process to the p53 tetramer entity with a process connector Step 5 Change the Kinetic Style of the process to Mass and set c1 Stoichiometry to 4 Leave the default values for the other parameters of the process Coefficient 0 01 Coefficient2 0 1 and c2 Stoichiometry 1 po3aimonameri a2 mi pl ct 02 pos tetramer _m2 10 m1 0 0010 i Figure 32 With the model as in Figure 32 the speed of the tetramerization process depends on the value of p53 tetramer To make the process of tetramerization faster the Coefficient value should be increased To change tetramerization into the dimerization process the cl stoichiometry value should be changed to 2 More details on parameters related t
93. rights reserved Page 50 of 100 x Process Figure Biological Mame C50 Class ProcessBiologica T Type Continuous Firing Style and Activity Gale style Kinetic Style Delay Priority Save Log Show Element Show Mame Show Speed Delay Show Firing Style References Comments Coefficient 1 Coefficient tandard deviation Figure 56 The major process parameters displayed in the Process Sheet are Type Firing Style Activity and Kinetic Style The Type property defines whether the process is discrete continuous or generic and determines its calculation style Calc Style field One can choose between two Firing Style settings and and or For a process with the default and firing style if all input process connectors and association connectors of the process are activated and no input inhibitory connector of the process is activated the process is enabled For a process with the or firing style the process is enabled if at least one of Association or Inhibitory input connector fulfills 1ts condition and all input process connectors fulfill their respective conditions The Kinetic Style defines the method for calculating the values of connected entities The property can have the following options custom connector custom mass stochastic mass or connector rate Each style requires a specific set of parameters that you need to enter at the bottom part of the Element settings dialo
94. rsions Until version 3 0 Cell Illustrator was offered as desktop application Since version 4 0 Cell Illustrator is available as Java Web Start application CI Online CI Online can be run directly from the website without the need to install software on local computer CI Online is a client server solution that enables the user to create manipulate and simulate bio pathway models locally within CI workspace and to store and organize them on the remote CI Online Server Additionally CI Online Server gives the user convenient access to libraries and databases of bio pathway models 2 2 Run Cell Illustrator Online CIO In order to run Cell Illustrator correctly Java version j2sdk 1 5 0 or higher needs to be installed Java Runtime Environments JREs are available for download at http java sun com downloads To run Cell Illustrator Online Open the Cell Illustrator Online website at https cionline hgc jp in your web browser E https cionline hac jp ci OA _ a iilimstrator Oniiine t0 Gell Ti ustrator Registration Cell iilustrator Player Java Web Start is required to run CIO Download the latest Java Runtime Enviroment JRE which includes Java Web Start and install it on your PC Get your login account through the registration and try CIO If you are already registered you can check your license status Copyright 2007 by GNI Ltd and The University of Tokyo Register to obtain the CI Onli
95. s model the values of el and e2 increase and decrease by the same amount 2 at each time interval because their value depends on the add value of the process p1 However modeling differing increase and decrease in the value of connected entities is sometimes necessary e g when a monomer becomes a dimer or a trimer and vice versa To deal with this type of process set the process kinetic style to connector custom and then define the value changes of the connected entities separately 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 18 of 100 me 9 a i F J Figure 6 The example in Figure 6 demonstrates the use of the connector custom kinetic style The difference between the previous example and the model in Figure 6 is as follows 1 The add parameters of connector cl and c2 are 2 and 1 respectively At time 0 the value of el is 10 and the threshold of cl is 1 and the process p1 is enabled After the delay of pl i e 1 the process is executed The add value of cl and c2 are 2 and 1 respectively Thus the value of e1 and e2 becomes 8 and 1 at time 1 respectively At time 4 the value of el and e2 are 2 and 4 respectively At that time the process p1 is not enabled and el and e2 reach a stable state The results of the simulation are presented below Figure 7 x Copy To Clipboard Figure 7 3 8 Continuous Ele
96. specially useful when importing models from the BioPACS database or converting models from other formats Graph Layout E xX E EJ py E all selected i others GEM Circle Import Export Close Sugiyama Figure 70 The Graph Layout dialog has several options for placing elements on the canvas such as Annealing Layout Circle Layout and others See the CI Reference Manual for details 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 63 of 100 6 Simulation 6 1 Simulation Concept The simulation architecture of Cell Illustrator is based on the Hybrid Petri net with extensions HFPNe theory 10 and uses the concepts of entities connectors and processes With discrete elements you can execute basic logic based simulations You can also execute differential equation based simulations with continuous elements and combine discrete and continuous elements in more complex models The most advanced simulations would employ generic elements and scripting For a short introduction to the HFPNe theory see Section 3 It is easy to prepare a basic simulation just by drawing biological pathways on the canvas A simple model can be then turned gradually into a more advanced one by adding detailed information to HFPNe elements in the biological pathways e g kinetics data and detailed regulation rules To build a model quickly d
97. st as one element The grouped elements can be ungrouped with the Ungroup operation Again there are three ways to ungroup elements via the popup menu in the canvas Element Ungroup in the main menu and the Ungroup icon in the Edit toolbar Cell Illustrator 1 5 DS Cell Ilustrator 1 5 DORI File Edit Element Simulation View Dialog Window Help File Edit Element Simulation View Dialog Window Help ER OTO pe l i ANRA S ESR e mk OODE L i 141411 2 2 Qe wri ER e E a B New Canvas 1 BEE a a E E ae oa ES R cs 2 a eal a E x a x oe oe 2 2 Q Teal lt Ti mj 1 e1 _ m1 e2 _ m2 r z if o1 si 6 x al ke 00 Ro A EE E oe te Boi gt gt D D rT il il Tort s a Zire dle gt _ Change default grid size done 2 202222222222 totalfree127MB79ME OOOO Groupthe selected elements dome totaldfree 127MB D4ME Figure 86 8 2 Hierarchization The Import options in the File menu and the canvas popup menu offer several ways to reuse your existing models as building blocks for creating a more complex model Import Model copies elements from a selected file into the model currently edited in the canvas Import As Group does the same with the following exception the imported elements are enclosed into one group element after the import operation Finally Import As Object inserts the mo
98. ting 8 152 163 Matsuno H Tanaka Y Aoshima H Doi A Matsui M Miyano S 2003 Biopathways representation and simulation on hybrid functional Petri net In Silico Biol 3 389 404 http www bioinfo de isb toc_vol_03 html Nagasaki M 2004 A Platform for Biopathway Modeling Simulation and Recreating Biopathway Databases Towards Simulation Ph D Thesis The University of Tokyo http genomicobject net pub nagasaki_phd pdf Nagasaki M Doi A Matsuno H Miyano S 2003 Recreating biopathway databases towards simulation In Computational Methods in Systems Biology Volume 2602 of Lecture Notes in Computer Science Springer Verlag pp 191 192 Nagasaki M Doi A Matsuno H Miyano S 2004 A versatile Petri net based architecture for modeling and simulation of complex biological processes Genome Informatics 15 1 180 197 Nagasaki M Doi A Matsuno H Miyano S 2004 Genomic Object Net I A platform for modeling and simulating biopathways Applied Bioinformatics 2 181 184 Nagasaki M Doi A Matsuno H Miyano S 2004 Integrating biopathway databases for large scale modeling and simulation In The Second Asia Pacific Bioinformatics Conference Volume 29 of Conferences in Research and Practice in Information Technology Australian Computer Society pp 43 52 Nagasaki M Doi A Matsuno H Miyano S 2005 Bioinformatics Technologies Chen Yi Ping Phoebe Ed Springer Press 179 243 ISBN 3540208739 Note Genomic Object Net is the resear
99. tion parameters or making corrections to the model itself The Simulation History dialog allows you to view and manage all the simulation log files CILs that have been saved for the given model file Ifthe simulation results are logged 1 e at least one entity or process is selected for logging see Simulation Log each simulation run is saved into a separate CI log file CIL along with the model representation at the time the simulation was done A log file has an extension ci and it is created in the folder project name output where project name is the name of the Cell Illustrator model file For example logs for the Cell Illustrator Apoptosis gon sample project are kept in the Apoptosis output subfolder in the folder where the Apoptosis gon file 1s located All the saved simulation logs are accessible in the Simulation History dialog box Simulation History xj Files simt cil interrupted an 5 2005 localhost the user sim2 cil finished Wan 13 200 localhost zima cil finished J ithe Weer note 2 Figure 79 This dialog allows you for viewing and comparing the properties of each simulation run The properties Author and Notes can be edited You can also manage the list of logs by removing unwanted CIL files from the list or adding CIL files that have been calculated on different computers by other users With the Replay button you can open the selected log files for replay in the Cell Illustrator Player app
100. tion runs will give different results due to the stochastic nature of the model 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 33 of 100 x Copy To Clipboard 100 t name _ cater pSa monamer ll Figure 35 4 7 Custom Tetramerization Tetramerization can also be defined by a custom formula In the example in Section 4 5 the process speed depends on the p53 monomer in a linear way but one can change it into a quadratic function by writing a script The example in 4 5 could be modified as follows Step 6 Change the Kinetic Style of the process to Connector Rate and set the Rate parameter as 0 001 m1 m1 Step 7 Make sure that other parameters are the same as in the previous Tetramerization example 1 e cl Stoichiometry is 4 pSSimonomer mi pi pSstetramer m2 C2 10 0 0 001 m1 m1 T Figure 36 In the model in Figure 36 the speed of the tetramerization process is proportional to the square of p53 monomer The results of the simulation are presented below Figure 37 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 34 of 100 x Copy To Clipboard name cor psaltetramer v MEN pSa monamer 7 ME Figure 37 4 8 Separation In a cell binding proteins are sometimes separated with proteins A model can be cr
101. ules binding to DNA ON OFF state of a gene expression etc Discrete Process A biological reaction that converts quantities in discrete entities into quantities in other discrete entities A switching mechanism can be also handled as a discrete process which receives some signals and sends out signals for a pathway Continuous Entity A biological entity like a concentration of a protein enzyme or ion etc the quantity of which can be represented as a real number Concentration Quantity System status representing quantity or concentration of proteins enzymes and ions Page 11 of 100 Continuous Transition A continuous transition is a continuous process consuming quantities of its input places and producing quantities into its output places It defines a continuous flow that is specified by a speed of firing v Continuous Process A biological process or a reaction like transcription translation or enzymatic reaction which consumes some entities inputs quantity and produces others output according to a rate formula Generic Entity Universal Place A universal place can handle a list of variables of various types double Boolean string etc A biological chemical physical entity that is not pre defined May contain more data types like string that is used to represent DNA sequence Universal transition Generic Process A transition for a uni
102. un can be logged and saved into CI log files CILs refer to Simulation Log Saving log files gives you the following possibilities e View and replay a simulation step by step in CI Player application refer to Simulation Replay e Order and systematically analyze the results for a set of similar simulations refer to Simulation History Creating a log file and replaying it in CI Player is recommended in the following cases e When the simulation may take a large amount of time e g for a complex model and accurate simulations e When you run a set of simulations for the same model and the simulation results may differ from one simulation run to another for example For a model with random factors or for a model with changed parameters 2002 2008 GNI Ltd and Human Genome Center Institute of Medical Science The University of Tokyo All rights reserved Page 65 of 100 6 2 How to Set Up and Run Simulation 6 2 1 Simulation Settings The next step after creating a model is to set up simulation process parameters The Simulation Settings dialog has the following options that control the simulation flow Sampling Interval Plot Update Interval Log Update Interval Simulation Time Continuous and Discrete Weak Firing Firing Accuracy and Simulation Speed Figure 74 During simulation the model changes its state in a certain time sequence In Cell Illustrator stimulation time is measured in virtual time units called Petri
103. versal place A A complicated biological chemical universal transition allows the arcs physical process e g a translation connecting with this transition to use their process functions firing speeds Arc Connector Process Connector Connects input entities to a process and process to output entities Its weight parameter specifies an activation threshold Process Connector allows quantities flow in the model Normal Arc A directed arc Connects a transition to a place or connects a place to a transition An arc has a weight Inhibitory Arc Inhibitory Connector Prevents a process from being activated Inhibitory Connector facilitates the modeling of the inhibition competition process Connects a place to a transition An inhibitory arc disables a transition If the quantity in its source place is greater than the weight of the arc the transition cannot fire Association Connector Connects entities without causing Test Arc concentration change Defines which Connects a place to a transition A test entities are associated with each other arc does not consume any quantity of the but not as inputs or outputs Association source place by firing Connector allows the modeling of enzymatic reactions as well as other catalytic processes 3 2 Entities Entities are abstract elements that can represent any type of biological concepts like mRNAs DNA proteins ligands and compounds They can also represent ce
104. ycle TCA cycle Pertose phosphate pathway Pentose and glucuronate interconversions Fructose and mannose metabolism Galactose metabolism Ascorbate and aldarate metabolism Starch and sucrose metabolisrn Aminosugars metabolism Nucleotide sugars metabolism Pyruvate metabolism Glyoxylate and dicarboxylate metabolism Propanoate metabolism Butanoate metabolism C5 Branched dibasic acid metabolism Inositol metabolism Inositol phosphate metabolism MI EEY e VS el aol ome oly v CERA ml ES ER EAA Add Inhibitor Add Effector Add Cofactor View Website Download Cancel Figure 89 Importing KEGG maps with BioPACS 8 5 Import of CellML or SBML models CellML www cellml org and SBML www sbml org are widely used XML file formats for representing mathematical relationships in biological models In Cell Illustrator you can import models from CellML SBML files into the CI workspace During the import operations the models are converted to CI models and can be edited analyzed simulated saved as any other models SBML CellML models differ from the CI ones E g SBML uses differential equations for representing relationships between entities while CI use the Petri Net model Therefore the conversion process has the following limitations o notall SBML models can be imported to CI o the simulation results might differ for a SBML model imported to CI It is recom

Cell Illustrator User Guide

Contents

Download Pdf Manuals

Related Search

Related Contents