dc.contributor.author
Li, Jian
dc.date.accessioned
2018-06-07T18:50:26Z
dc.date.available
2013-11-07T14:37:08.685Z
dc.identifier.uri
https://refubium.fu-berlin.de/handle/fub188/5487
dc.identifier.uri
http://dx.doi.org/10.17169/refubium-9686
dc.description
Computational Modeling of Cancer-Associated Cellular Systems 2
Dedication..........................................................................................................3
Acknowledgements............................................................................................4
English
Abstract................................................................................................
6 German
Abstract................................................................................................7
List of Tables 14 List of Figures 15 Abbreviations 16 1\. Introduction 20 1.1
Tumorigenesis 20 1.1.1 Deregulation of Cellular Regulatory Circuits 21 1.1.2
Personalized Medicine 23 1.2 MicroRNA Regulation Regarding Cancer research 25
1.2.1 MiRNA & Cancer Research 26 1.2.1 MiRNA & EGFR Signaling Pathway 27 1.3
Human Signaling Model 28 1.4 Human Metabolic Model 30 1.5 Gene-regulatory
Network 33 1.6 Systems Biology & Software 34 1.6.1 CellDesigner 36 1.6.2
Copasi 37 1.6.3 PySCeS 39 1.6.4 Virtual Cell 40 1.6.5 PyBioS2 42 1.7 Objective
and Focus of this Thesis 44 2\. Results 44 2.1 Establishment of an Integrated
miRNA-EGFR Signaling Pathway Model 44 2.2 Validation of the Predictive Ability
of the miRNA-EGFR Model 48 2.3 Simulation of Functional Ability of mir-192 and
mir-181 50 2.4 Evaluation of the “One Hit – Multiple Targets” Concept using
miRNA Modeling 54 2.5 Establishment of A Non-Steroidal Anti-Inflammatory Drug
Model 59 2.6 Integration of Cancer Hallmarks into NSAID Model and Its
Validation 62 2.7 Analysis of COX Based Synthetic Lethality for Breast, Colon
and Lung Tumor 66 2.8 The challenges of Construction of A Comprehensive
Signaling Model in Human 71 2.9 Components Identification of the Human
Signaling (HS) Model 71 2.10 Reassembling Model Components Focused on the
Systems-biological Properties 73 2.11 Analysis of Dynamic Behavior of the
Biological System underlying the Human Signaling Model 80 2.12 Establishment
of a Tumorigenesis Calculation Formula 90 2.13 Utilization of miRNA-expression
Data to Discover and Detect the Biomarkers of Individual Cancer Cell Lines 92
2.14 Reconstruction of the Human Metabolic Model and Extension with the miRNA-
Regulation Network. 94 2.15 Features and Functionalities of PyBioS2 96 2.15.1
Multiple Database-Connection 97 2.15.2 Data Integration 98 2.15.3 Graphical
Interface 100 2.15.4 Model Architecture 101 2.15.5 Semi-Automatic Definition
of Reaction 102 2.15.6 ODE System 104 2.15.7 Petri net 105 2.15.8 Utilization
of Gene-Expression Data 107 2.15.9 Model Analysis Approaches 108 2.15.10 A
Case Study: Comparison of Petri-net simulation strategy and ODE simulation
strategy. 110 3\. Materials and Methods 110 3.1 Petri net Extension 110 3.2
Analysis of Significant Changes of Model Components due to the Effect of anti-
miRNAs 115 3.3 Model Construction 115 3.4 Model Parameterization 116 3.5 Model
Initialization with Gene-expression or miRNA-expression Data and Model
Simulation Procedure 116 3.6 Comparison of Predicted Proteome and Experimental
Proteome Data 117 3.7 Establishment of A Drug Database 118 3.8 Weight-Factor
Estimation for the Tumorigenesis Calculation Formula 119 3.9 Cancer Cell Lines
from NCI-60 Stduy and from Cancer Genome Atlas 119 3.10 Sensitivity Score
Definition 120 3.11 The implementation of cancer- and process-hallmarks in the
human signaling model. 120 3.12 Mathematical Implementation of COX-isoform
Specific siRNA and NS-398 Drug Effect 122 4\. Discussion 122 4.1 The miRNA-
EGFR (ME) Model 122 4.2 The Non-Steroidal Anti-Inflammatory Drug (NSAID) Model
127 4.3 The Human Signaling (HS) Model 128 4.4 The Human Metabolic Model
(HMMA) 135 4.5 PyBioS 2 137 4.6 Conclusions and Future Outlook 139 Reference
142 Publication 156 A. Supplemental Information 159 A.1 miRNA-EGFR model 159
A.2 MicroRNA target validation references for miRNA-EGFR model 159 A.3 Pathway
references for human signaling model 159 A.4 MicroRNA target validation
references for human signaling model 159 A.5 Feedback loops in the human
signaling model 159 A.6 PyBioS 2 Tutorial (step by step) 159 A.7 Semi-
automatic reaction definition 173 A.8 Gefitinib, Imatinib and Temsirolimus
Drug Association Constant 175 A.9 MicroRNA target validation references for
human metabolic model (HMMA) 176 A.10 The Human Metabolic miRNA Model (HMMA)
176 A.11 The Human Signaling Model 176 A.12 The Non-Steroidal Anti-
Inflammatory Drug (NSAID) Model 176 B. Supplemental Figure 177 B.1 The
connectivity-degree distribution of model components with a linear regression
line. 177 B.2 The correlation plot between experimental proteom data and
predicted proteome data from the u-2 os cell line (attached in CD). 178 B.3
The correlation plot between experimental proteom data and transcriptome data
from the u-2 os cell line (attached in CD). 178 B.4 The correlation plot
between experimental proteom data and predicted proteome data from the A-431
cell line (attached in CD). 179 B.5 The correlation plot between experimental
proteom data and transcriptome data from the A-431 cell line (attached in CD).
179 B.6 The correlation plot between experimental proteom data and predicted
proteome data from the U-251 MG cell line (attached in CD). 180 B.7 The
correlation plot between experimental proteom data and transcriptome data from
the U-251 MG cell line (attached in CD). 180 B.8 The in-silico simulation
experiments of human signaling pathways. 181 B.9 Comparison of ODE simulation
and Petri net simulation based on a simple network. 182 B.10 The cancer cell
lines from Cancer Genome Atlas.....................................183 C.
Supplemental Table 183 C.1 Pathway description 183 C.2 Pathway, key component,
transcriptional target and major involved microRNA 198 C.3 Pathway, Ligand,
Receptor, downstream targets and crosstalk pathways 198 C.4 The simulation
data of 100 specific miRNA Inhibitors 210 C.5 The simulation data of AKT and
MEK inhibitor 210 C.6 The simulation data of single ligand-activations in the
human signaling model. 210 C.7 The simulation data of 100 specific miRNA
Inhibitors 211 C.8 The comparison result of the simulation data with the data
from Nagaraj's Study 211 C.9 The drug response prediction data 211 List of
Tables PyBioS2_Database_Connection 43 miRNA-EGFR_Model_Component_Summary 45
Top15_anti-miRNA_Inhibitor 58 NSAID_Model_Component_Summary 61
Human_Signaling_Model_Component_Summary 73 Statistics_Feedback_Loops 74
Four_Simulation_Conditions 75 Statistics_Bow_Tile_Structure 78
Hallmarks_Mechanism_Related_Pathways 82 Biomarker_Identification_Result 94
HMMA_Model_Component_Summary 95 Function_Comparison_Systemsbiology_Softwares
97 Semi-Automatic_Reaction_Type 103 Network_Analysis_Methods 108
Component_Concentrations_at_TimePoint_0 114
Component_Concentrations_at_TimePoint_1 114 Kinetic_Parameter_Summary 115
Cancer_Hallmarks_Implementation 121
Mathematical_Implementation_Therapeutic_Intervention 122
Gefitinib_Imatinib_Temsirolimus_Drug_Association_Constant 175
Pathway_Description 183 Pathway_Ligand_Component_Information 198 List of
Figures CellDesigner Annotation Interface 37 Copasi User Interface 38 PySCeS
Command Line and Output Graph 40 Visualization of Virtual Cell Interface 41
PyBioS2 User Interface and Model Visualization 43 miRNA-EGFR Model Network
Overview 46 Comparison of miRNA-EGFR Model Predictions with Experimental
Results. 49 Modeling of mir-192 and mir-181c Effects on the EGFR Signaling
Pathway 51 Modeling the Individual Effect of 100 anti-miRNAs. 56 Histogram of
miRNA/target Relationship 58 The Network Graph of COX Pathway 61 Simplified
Visualization of COX-2 Inhibitions within NSAID Model Network 64 Effect of the
COX-based Combined Inhibition on Cancer Hallmarks 67 Simplified Human
Signaling Model Overview 72 Schematic Representation of the Concept for
Functional Redundancy in the Human Signaling Model 76 Heatmap of log2-ratios
from the Concentrations of Downstream Components in this Network 76 The Bow-
Tile Structure 77 Functional Modularity Regarding the Relationship Between
Pathways and Hallmarks 79 Ligand-Activation-Test 84 miRNA Regulation Regarding
the Effect of Ligand Activations 84 Workflow of in-silico Simulation Pipeline
86 Signal Propagation Strategy for Inheriting Dynamic Properties of the
Underlying Biological System 87 Visualization of the Correlation Measurements
A 88 Visualization of the Correlation Measurements B 88 Drug Response
Prediction of NCI-60 Cancer Cell Line A 93 Drug Response Prediction of NCI-60
Cancer Cell Line B 93 Schematic Overview of PyBioS2 Functions 98 Diagram of
PyBioS2 Internal Database 99 Navigation-Frame and Content-Frame of PyBioS2 101
Internal Structure of a Model in PyBioS2 102 Visualization of Model Network
under ODE System Module 104 Visualization of Model Network under Petri Nets
Module 107 Transition Fire of a Petri Net 112 Drug Database Diagram 118
dc.description.abstract
Cancer is the consequence of disordered cellular systems. To date, more than
hundred distinct types of cancer and subtypes of tumor have been identified in
the human body. This complexity does not only provoke a number of questions
such as the origin of cancer and its medical treatment, but also poses a high
level of challenge to battle against them. The experience gained by diverse
cancer-research studies suggests that all differences among traditional cancer
and tumor types can be reduced at the molecular level. Thus, it is essential
to understand the interplay and relation between oncogenes and tumor
suppressors within the cancerous cellular system, which could help to identify
the core function of molecular pathways for effective therapeutic
intervention. This study tried to realize an important system biological
concept that integrates modeling of different types of biological networks
including signaling-, metabolic-, gene-regulatory-, and miRNA-regulation-
network from the whole organism and generate responsive computational
molecular models to describe the behavior of cells (or cancer cells). The
models created during this thesis are demonstrated with application-examples
to be able to incorporate the experimental and observational inquiry and
receive the individual patient information to reflect/predict the dynamic
behavior of the underlying cellular systems. This study also tried to
elucidate that models of this kind can be utilized for early tumor detection,
precise diagnose of the pathological state, the identification of anti-
tumor/cancer drug and drug combination effect. Therefore, models of this kind
might be able to form a “Virtual Patient” model to propose, verify and predict
new therapeutic strategies towards personalized medicine. In addition, a
systemsbiological software has been developed during this study. Its novel
properties including its internal modeling database can be used to overcome
some current limitations of systemsbiology research.
de
dc.description.abstract
Krebs ist eine Art von Krankheit, die durch funktionale Störungen innerhalb
zellulärer Systeme verursacht wird. Bei Menschen und Tieren bezeichnet Krebs
maligne Tumoren, die unkontrolliertes Wachstum zeigen, was auf den Verlust der
Proliferationskontrolle zurückzuführen ist. Krebs kann aus allen sich
teilenden Zelltypen entstehen, sodass bis heute mehr als 100 verschiedene
Krebserkrankungen bekannt sind. Gegenstand der Krebsforschung heutzutage ist
auf der einen Seite die Aufklärung der Krebs-Entstehung und auf der anderen
Seite die Entwicklung von neuen bzw. verbesserten medizinischen
Behandlungmöglichkeiten für Krebs. Die aus diversen Studien gewonnenen
Erkenntnisse erwecken den Eindruck, dass alle Unterschiede zwischen
traditionellen Krebs- und Tumortypen auf die molekulare Ebene zurückgeführt
werden können. Daher ist es wichtig, das Zusammenspiel zwischen Onkogenen und
Tumorsuppressorgenen in den molekularen Prozessen die zur einzelnen Krebszelle
führen, noch genauer zu verstehen, um Krebs auf molekularer Ebene effektiver
zu bekämpfen und schließlich zu besiegen. Diese Arbeit versucht, ein wichtiges
systembiologisches Konzept zu verwirklichen, indem ein integratives
mathematisches Computer-Modell entwickelt wurde, das unterschiedliche
biologische Netzwerke (Signal-, metabolische Netzwerke, genregulatorische
Netzwerke und miRNA Regulationsnetzwerke) beinhaltet. Dieses integrative
Modell besteht aus verschiedene Sub-modelle, die während der Arbeit aufgebaut
und validiert worden sind. Die Arbeit hat auch aufgeführt wie man Modelle
dieser Art mit Anwendung/Integration von der genetischen Information der
individuellen Zelllinien and experimentellen Daten verwenden kann, um das
dynamische Verhalten von Krebszellen aufzuklären. Diese Arbeit versucht auch,
zu erklären, dass Modelle dieser Art auch für medizinische Anwendungen, z.B.
die Früherkennung von Tumoren, die Identifizierung der Wirkstoffe und Diagnose
der Wirkung von Anti-Krebs-Medikamenten oder Anti-Krebs-
Arzneimittelkombinationen, einsetzbar sind. Daher könnten diese Modelle in der
Lage sein, als ein "Virtueller Patient" zu funktionieren, und neue
therapeutische Strategien zur personalisierten Medizin vorschlagen, überprüfen
und vorhersagen. Zusätzlich wurde eine systemsbiologische Software während der
Arbeit entwickelt, derer vorteilhafte Eigenschaften, inklusive deren internale
Datenbank, die dazu gedient werden kann, Beiträge zur systemsbiologische
Forschung zu leisten.
de
dc.rights.uri
http://www.fu-berlin.de/sites/refubium/rechtliches/Nutzungsbedingungen
dc.subject
systems modeling
dc.subject
signaling pathways
dc.subject
miRNA-regulation
dc.subject
biomarker identification
dc.subject
drug reponse prediction
dc.subject.ddc
500 Naturwissenschaften und Mathematik::570 Biowissenschaften; Biologie::570 Biowissenschaften; Biologie
dc.title
Computational Modeling of Cancer-Associated Cellular Systems
dc.contributor.contact
jli@mi.fu-berlin.de
dc.contributor.firstReferee
Prof. Dr. Burghardt Wittig
dc.contributor.furtherReferee
Prof. Dr. Peter N. Robinson
dc.date.accepted
2013-10-30
dc.identifier.urn
urn:nbn:de:kobv:188-fudissthesis000000095449-8
dc.title.translated
Computergestützte Modellierung von Krebs-assoziierten zellulären Systemen
de
refubium.affiliation
Biologie, Chemie, Pharmazie
de
refubium.mycore.fudocsId
FUDISS_thesis_000000095449
refubium.note.author
Zum Betrachten der Supplement-Dateien diese erst lokal speichern, dann öffnen,
sonst evtl. Fehlermeldung
refubium.mycore.derivateId
FUDISS_derivate_000000014345
refubium.mycore.derivateId
FUDISS_derivate_000000014346
refubium.mycore.derivateId
FUDISS_derivate_000000014347
refubium.mycore.derivateId
FUDISS_derivate_000000014358
dcterms.accessRights.dnb
free
dcterms.accessRights.openaire
open access