dc.contributor.author
Reimers, Arne C.
dc.date.accessioned
2018-06-07T23:52:36Z
dc.date.available
2014-09-16T10:37:32.329Z
dc.identifier.uri
https://refubium.fu-berlin.de/handle/fub188/11116
dc.identifier.uri
http://dx.doi.org/10.17169/refubium-15314
dc.description
Abstract Zusammenfassung Acknowledgments 1 Introduction 1.1 Metabolic Networks
1.2 Steady-State Assumption 1.3 Thermodynamic Constraints 1.4 Constraint-Based
Analysis 1.4.1 Does the organism grow under certain conditions? 1.4.2 How does
the organism grow? 1.4.3 How can a growth behavior be enforced? 1.5
Mathematical Context 1.6 Structure of this Thesis 2 Feasible Pathways 2.1
Basic Concepts and Notation 2.1.1 Directed Reactions and Equivalence 2.1.2
Flux Coupling 2.2 Steady-State Assumption 2.2.1 Average Fluxes are Steady-
State 2.2.2 Concentrations for Average Fluxes 2.3 Polyhedral Flux Spaces 2.4
Elementary Modes 2.5 Matroid Theory 2.5.1 Notation and Definition 2.5.2
Matroids from Metabolic Networks 2.5.3 Matroid Connectivity 2.6 Thermodynamic
Constraints 2.6.1 Chemical Potentials 2.6.2 Without Bounds on Chemical
Potentials 2.6.3 With Bounds on Chemical Potentials 3 Computational Complexity
3.1 Introduction 3.2 Results 3.3 Reductions 3.3.1 Trivial Reductions 3.3.2
1x2y3a4a is Equivalent to 1x2y3b4a for x ∈ {a, b}, y ∈ {b, c} 3.4 Problems in
P 3.4.1 1a2c3a4a is in P 3.4.2 1b2x3a4a is in P 3.4.3 1b2b3b4a is in P 3.4.4
1x2y3c4a is in P 3.4.5 1x2c3b4a is in P 3.5 NP-hard problems 3.5.1 1a2a3a4x is
NP-hard 3.5.2 1x2a3b4a and 1x2y3b4b are NP-hard 3.5.3 1x2y3c4b and 1x2y3a4b
are NP-hard 3.6 Problems with Unknown Complexity 4 Flux Optimization 4.1 Flux
Balance Analysis 4.2 Thermodynamic Constraints 4.3 Mixed Integer Linear
Programming 4.4 Implicit Representation of the Potential Space & Constraint
Programming 4.4.1 Practical Implementation 4.4.2 Coupled Reactions and
Generalized Infeasible Sets 4.4.3 Partitioning the Flux Space 4.5 Heuristics
4.5.1 Cycle Subtraction 4.5.2 First Directions, then Fluxes 4.6 Conclusion 5
Potential Optimization 5.1 Strict Inequalities 5.1.1 Strict Inequalities in
Linear Programming 5.1.2 Strict Inequalities in Mixed Integer Linear
Programming 5.1.3 Integration into MILP-solver 5.1.4 Application to
Thermodynamic Constraints in Metabolic Networks 5.2 Improving Bounds on
Potentials 5.2.1 Introduction: the Graphic Case 5.2.2 The Flow Condition 5.2.3
Updating Bounds 5.2.4 Application on a Genome-Scale Network of E. coli 5.2.5
Dependence of Uncertainties in Equilibrium Constants 6 Modules in Metabolic
Networks 6.1 Definition 6.2 Properties of (Linear) k-Modules 6.2.1 Restriction
to Linear Vector Spaces 6.2.2 Matroid Theory for k-Modules 6.2.3 Finding
k-Modules 6.3 Flux modules (0-modules) 6.3.1 Uniqueness of the Decomposition
6.3.2 Decomposition Theorem for EFMs 6.3.3 Finding Flux Modules 6.3.4
Comparison of the Methods 6.3.5 Computed Modules in Genome-Scale Metabolic
Networks 6.3.6 Visualizing the Interplay of Flux Modules 6.3.7 Conclusion 6.4
Decomposition Theorem for Linear 1-Modules 6.4.1 Computation of Linear
1-Modules 6.4.2 Linear 1-Modules in Practice 6.5 Decomposition Theorem for
k-Modules 6.5.1 Elementary Flux Mode Enumeration and Vertex Enumeration 6.5.2
Branch-Width 6.5.3 Decomposition Theorem for Vertex Enumeration 6.6 Conclusion
7 Sublinear Growth & Flux Forcing Reactions 7.1 Motivation 7.2 Flux-Forcing
Reactions can Explain the Effect 7.3 Sets of Diverting Reactions 7.3.1
Diverting Set of Type 1 7.3.2 Diverting Set of Type 2 7.3.3 Finding Flux
Forcing Coefficients for Diverting Sets of Type 2 7.3.4 Thermodynamically
Constrained P 7.4 Multi-Parametric Thermodynamically Constrained Flux Balance
Analysis (mpTFBA) 7.4.1 The Algorithm by Dua and Pistikopoulos 7.4.2
Adaptation to Thermodynamically Constrained FBA 7.4.3 Finding Cut Sets for
Blocking Internal Cycles 7.5 Tight Integration of Bilevel Programming into
Parametric Programming 7.5.1 Min-Max Problems 7.6 Implementation 7.7 Results
7.8 Discussion 8 Flux Coupling Analysis with Thermodynamic Constraints 8.1
Introduction 8.2 Lattice Theory 8.2.1 FCA for arbitrary P 8.2.2 Algorithm for
FCA in P 8.3 FCA in T 8.4 Implementation 8.5 Minimal Representation of
Coupling Data 8.5.1 Minimal Extensions 8.5.2 Computation of Minimal Extensions
8.6 Discussion 8.6.1 Theoretical Differences 8.6.2 Practical Differences 8.7
Conclusion 9 Sampling the Thermodynamically Constrained Flux Space 9.1
Background 9.2 Theoretical Obstructions to Sampling 9.3 Practical Obstructions
to Sampling 9.3.1 Method 9.3.2 Results 9.4 Discussion 9.5 Conclusion A
Computational Results on Flux Modules A.1 Flux Modules A.1.1 Summary on E.
coli iJR904 grown on L-tryptophan A.1.2 Summary on E. coli iJR904 grown on
L-Threonine A.1.3 Summary on E. coli iJR904 grown on glucose A.1.4 Summary on
E. coli iJR904 grown on L-Arginine A.1.5 Summary on E. coli iJR904 grown on
citrate A.1.6 Summary on E. coli iJR904 grown on fumarate A.1.7 Summary on E.
coli iJR904 grown on L-glutamine A.1.8 Summary on E. coli iJR904 grown on
Lactose A.1.9 Summary on E. coli iJR904 grown on L-malate A.1.10 Summary on E.
coli iAF1260 grown on glucose, aerobic A.1.11 Summary on E. coli iAF1260 grown
on glucose, anaerobic A.1.12 Summary on E. coli iAF1260 grown on L-Threonine
A.1.13 Summary on H. pylori iIT341 A.1.14 Summary on M. barkeri iAF692 grown
on methanol A.1.15 Summary on M. tuberculosis iNJ661 A.1.16 Summary on S.
aureus iSB619 A.1.17 Summary on S. cerevisiae iND750 grown on D-glucose A.2
Optimal-Yield Elementary Flux Modes A.2.1 Elementary Flux Modes of E. coli
iJR904 grown on L-Arginine A.2.2 Elementary Flux Modes of E. coli iJR904 grown
on L-Threonine B Notation Bibliography Index Kurzzusammenfassung Curriculum
Vitae Declaration
dc.description.abstract
Biological experiments are time-consuming and expensive. Hence, computer-aided
experimental design is used more and more often to select those experiments
that are likely to yield new insights. In this work I consider variants of the
constraint based method Flux Balance Analysis. Constraints are used to exclude
biologically unrealistic phenotypes. Many constraints (for example flow
conservation) can be formulated using linear inequalities, which allows
efficient analysis using linear programming. Thermodynamic constraints are
used to include also energetic aspects. However, thermodynamic constraints are
not linear and induce a non-closed solution space. In this work I show that
this non-linearity of thermodynamic constraints often leads to NP-hard
decision problems. I show how this has consequences on the reliability of
sampling methods. However, I also present solution approaches that allow us to
solve optimization problems and qualitative analysis methods, like flux
coupling analysis with thermodynamic constraints efficiently in practice.
These insights I then use to develop a bi-level optimization method to analyze
a growth behavior of the green alga Chlamydomonas reinhardtii. Another area
covered by this work are flux modules. Based on a work by Kelk et al., I give
a mathematical definition of flux module and show that this definition
satisfies the desired properties. This definition also allows me to show
several decomposition theorems. These decomposition theorems simplify the
analysis of metabolic networks. Using matroid theory I show how modules can
efficiently be computed. With the definition of k-module, I also show a
decomposition theorem for general polyhedra using the concept of matroid
branch-width. With flux modules and algorithmic approaches to also include
complex constraints I present methods in this thesis that simplify and speed-
up the analysis of metabolic networks. This allows us to gain biological
insights faster and develop better methods for the production of bio-fuels in
bio-engineering and cancer therapies in medicine.
de
dc.description.abstract
Biologische Experimente sind zeitraubend und teuer. Deshalb werden Computer
immer häufiger benutzt um solche Experimente zu bestimmen, die am ehesten
erfolgreich sind und zu neuen Erkenntnissen führen. In dieser Arbeit betrachte
ich Varianten der Constraint-Basierten Methode Flussbalance-Analyse. Mittels
Nebenbedingungen werden biologisch unrealistische Verhaltensweisen
ausgeschlossen. Viele der Nebenbedingungen (z.B. Flusserhaltung) lassen sich
mit linearen Ungleichungen beschreiben, was effiziente Analyse mittels
linearer Programmierung ermöglicht. Thermodynamische Nebenbedingungen sorgen
dafür, dass auch energetische Aspekte berücksichtigt werden. Allerdings sind
diese Nebenbedingung nicht linear und induzieren zudem oft einen nicht
abgeschlossenen Lösungsraum. In dieser Arbeit zeige ich, dass diese nicht-
Linearität der thermodynamischen Nebenbedingungen häufig zu NP-schweren
Problemen führt, was sich z.B. auf die Verlässlichkeit von Samplingmethoden
auswirkt. Aber ich diskutiere auch Lösungsansätze, wie dennoch in der Praxis
Optimierungsprobleme und qualitative Analysemethoden, wie
Flusskopplungsanalse, mit thermodynamischen Nebenbedingungen effizient gelöst
werden können. Anwendung finden diese Erkenntnisse in der Analyse des Wachstum
der Grünalge C. reinhardtii, wo ich eine Methode zum Lösen von bilevel-
Optimierungsproblemen mit thermodynamischen Nebenbedingungen entwickle. Ein
weiteres Gebiet meiner Arbeit umfassen Flussmodule. Inspiriert von einer
Arbeit von Kelk et al. gebe ich eine mathematische Definition von Flussmodul
und zeige dass die Definition die erwünschten Eigenschaften erfüllt. Die
Definition erlaubt mir auch mehrere Zerlegungsätze zu zeigen, die die Analyse
metabolischer Netzwerke vereinfachen, und Module effizient mittels
Matroidtheorie zu finden. Mit der Definition von k-Modul zeige ich auch einen
Zerlegungssatz für die Eckenenumeration allgemeiner Polyeder und nutze hierfür
das Konzept der Branchweite in Matroiden. Mittels Flussmodulen und
algorithmischen Ansätzen um auch komplizierte Nebenbedingungen zu integrieren,
zeige ich in dieser Arbeit Methoden auf, die die Analyse metabolischer
Netzwerke vereinfachen und beschleunigen. Dadurch können biologische
Erkenntnisse schneller gewonnen werden und bessere Methoden in der
Biotechnologie zur Herstellung von Biotreibstoffen und in der Medizin für
Krebstherapien entwickelt werden.
de
dc.format.extent
XVI, 281 S.
dc.rights.uri
http://www.fu-berlin.de/sites/refubium/rechtliches/Nutzungsbedingungen
dc.subject
metabolic network
dc.subject
thermodynamic constraint
dc.subject
vertex enumeration
dc.subject
integer programming
dc.subject
bi-level optimization
dc.subject.ddc
500 Naturwissenschaften und Mathematik::510 Mathematik
dc.subject.ddc
500 Naturwissenschaften und Mathematik::570 Biowissenschaften; Biologie
dc.subject.ddc
000 Informatik, Informationswissenschaft, allgemeine Werke::000 Informatik, Wissen, Systeme::005 Computerprogrammierung, Programme, Daten
dc.title
Metabolic Networks, Thermodynamic Constraints, and Matroid Theory
dc.contributor.contact
arne.c.reimers@gmail.com
dc.contributor.firstReferee
Prof. Dr. Alexander Bockmayr
dc.contributor.furtherReferee
Prof. Dr. Leen Stougie
dc.date.accepted
2014-08-14
dc.identifier.urn
urn:nbn:de:kobv:188-fudissthesis000000097358-5
dc.title.translated
Metabolische Netzwerke, Thermodynamische Nebenbedingungen und Matroidtheorie
de
refubium.affiliation
Mathematik und Informatik
de
refubium.mycore.fudocsId
FUDISS_thesis_000000097358
refubium.mycore.derivateId
FUDISS_derivate_000000015787
dcterms.accessRights.dnb
free
dcterms.accessRights.openaire
open access