Lipid metabolism of Plasmodium falciparum

During its disease causing stage, the malarial parasite Plasmodium falcipaum replicates within human erythrocytes. Within one replication cycle one parasite can multiply up to 40-fold. This replication process demands an enormous supply of lipids, fueling the organization of new parasite membranes. Despite de-nove synthesis, lipids can be obtained from the blood serum as well as scavanged from the erythorcyte membrane (afterall its a parasite). The total lipid amount as well as the lipid composition changes during the parasites life cycle. It remains vague where the newly aquired lipids originate from. Describing lipid uptake, scavange and synthesis with a mathematical model could indicate the contributions of each of these processes to the observed change in lipid numbers and composition. An agent based model approach is used as emzymes participating in lipid metabolism are often promiscuous and the number of possible subtrates is vast.

METHODS

• agent based modeling
• stochastic modeling

PEOPLE

Jorin Diemer      Lasse Bonn

Large-scale economic model of growing budding yeast

 Building and analysing a Resource Balance Model of Baker's yeast.

METHODS

• Genome-scale metabolic models
• Constraint-based analysis/methods.
• Linear optimisation
• Cellular resource allocation

PEOPLE

Oliver Bodeit

Metabolism in cancer

Modeling the central carbon metabolism for three different colon cancer cell lines. The goal is to find post translational modifications (PTMs) which explain differences between the cell lines that cannot be explained by changes in enzyme concentrations alone. To do that, a model is fitted for the "wildtype" cell line and then this model is used to try to explain the data of other cell lines by using their enzyme concentrations. If these data are not explained well, PTMs are added to selected kinetics of the model to better describe the data.

METHODS

• R, dMod
• 13C carbon tracing
• pSIRM workflow
• metabolic flux analysis (MFA)

PEOPLE

Paula Martinell Garcia       Roman Rainer       Dr. Timo Lubitz

Modeling mitochondrial metabolism

Mitochondria are cellular organelles that harbour complex biochemical processes, and a purely experimental approach is insufficient to understand the intricate interactions of these biochemical networks. Thus, we aim to develop a kinetic model of mitochondrial metabolism suitable to study its complex interactions in a physiological and pathological context.

METHODS

Thermodynamic-kinetic ODE modeling

• parameter estimation
• Matlab
• D2D
• R

PEOPLE

Olufemi Bolaji

Agent based modeling lipid metabolism in yeast

Promiscuity of metabolic enzymes (using different substrates) can complicate the deterministic modeling of metabolic networks. These enzymes are especially prominent in lipid metabolism where different fatty acids and headgroups can build a vast pool of different substrate molecules. Firstly, classification of substrates is not possible, every molecule type has to be represented by its own species, leading to a large number of variables. Secondly, it is difficult to estimate the enzyme allocation, e.g. which and how many substrates are used by the enzymes. Therefore, we're working on a agent-based modeling approach in which every enzyme molecule acts as an agent and can select stochastically it's substrate molecules. 

METHODS

• agent based modeling
• Python
• stochastic modeling

PEOPLE

Jens Hahn       Dr. Katja Tummler       Dr. Josch Konstantin Pauling       Pascal Iversen

Metabolic flexibility in mycobacteria

 METHODS

• thermodynamic-kinetic ODE modeling
• parameter estimation (deterministic optimizers)
• metabolic control analysis
• time-resolved metabolomics (Sauer group, ETH Zürich
• quantitative proteomics (Aebersold group,  ETH Zürich)

PEOPLE

Dr. Katja Tummler

Modeling human liver and kidney metabolism

This project is part of the collaborative SysToxChip project that generally deals with liver and kidney responses to toxicity in order to design a microfluidic chip for the prediction of drug toxicity. To understand the metabolic connection between liver and kidney, that are the two major detoxificating organs in mammals, we wanted to analyze interorganic metabolic fluxes. However, the existing metabolic models for liver and kidney differed too much in detail, as well as in granularity. Therefore, we decided to expand the existing kidney model from Chang et al. 2010 and manually curate it in order to reach a minimal required detail for connecting it to the much larger metabolic liver model HepatoNet (Gille et al. 2010). Up to now 273 reactions and 187 metabolites have been added and for all included reactions the charge and atom balances have been checked. Currently, the model is tested for completeness and dead ends and in a next step the inclusion of a general drug metabolic pathway is planned. This project is supported by collaborative work within the SysToxChip consortium: Mrowka group, Universitätsklinikum Jena: development of a plasmid reporter system for toxicity in liver & kidney Wölfl group, Universität Heidelberg: experimental analysis of toxicity responses Andrade group, Universität Mainz: bioinformatic analysis of toxicogenomics database Kurtz group, Charité Berlin: differentiation of iPS Genomatix München: pathway reconstruction Chip Shop GmbH Jena: microfluidic chip development

METHODS

• constraint-based modeling
• flux balance analysis (FBA)
• extreme pathway analysis
• metabolic control analysis (MCA)
• manual curation

PEOPLE

Dr. Judith Wodke       Thomas Marschall       Laura Stumpf

Liver lipid metabolism

Non-alcoholic fatty liver disease (NAFLD) is thought to be one possible manifestation of the metabolic syndrome. Unravelling the causes and factors leading to the disease and its progression is the focus of the Liver Systems Medicine (LiSyM) project. Several studies show that one key aspect to the disease are changes in the hepatic lipid metabolic network, especially fatty acid composition and phospholipid remodeling. To understand this intertwined lipid machinery, its fluxes, and enzyme activities a computational model is being developed. A comprehensive set of genomics, proteomics, and lipidomics data is available ranging from mouse models on various diets to data of human patients suffering from NAFLD in different progression states. Utilizing these data to parameterize the computational lipid metabolic model and subsequent quantitative simulations will provide insights into the disease state-specific alterations and system dynamics in response.

METHODS

• metabolic network reconstruction
• network topology
• statistical data analysis
• stochastic agent-based modeling
• kinetic modeling
• lipidomics and lipidology

PEOPLE

Dr. Josch Konstantin Pauling

Thermodynamic modeling of metabolism

Metabolic reactions underlie thermodynamic constraints. When modeling metabolism, these constraints are often ignored due to simplifications. Within our project we aim at modeling the kinetics of metabolic reactions by enforcing thermodynamic constraints on the reactions. We developed an approach called parameter balancing, which is a kinetic parameter estimation within a Bayesian framework. It produces numeric values for unknown kinetic parameters on the basis of known kinetic parameters and prior distributions. This tool can be used online via a user interface for SBML models and SBtab parameter files.

METHODS

• Bayesian estimation
• Python
• web2py
• SBtab

PEOPLE

Dr. Timo Lubitz

Redox fluctuations

Metabolic oscillations in yeast cells have been shown to occur in both cycling an arrested cells. The duration and amplitude of the oscillations adjust to the cell's requirements and enviromental conditions. We are interested in modeling the way the activation of the cell cycle modifies the oscillations of ATP and NADH and how these oscillations behave in the absence of the division cycle. An ordinary differential equations (ODEs) model is constructed in order to simulate single cell dynamics. Experimental data on ATP levels in sigle cells will be generated to adjust the model.

METHODS

• fluorescence microscopy in single cells
• ordinary differential equation (ODE) modeling  

PEOPLE

Paula Martinell Garcia