Currently, the yeast cell model framework is a meta-simulation platform for ODE models. It comes with a graphical user interface. The simulation pipeline is driven by wizard steps where the user can load modules, modify values, simulate and test modules and organize the simulation output. The modules can be described by chemical reactions and/or as ODE equations. The model description language is SBML (level 3) or an in-house variant, developed by our group. The annotation of the species follows strictly the MIRIAM-Registry (Minimum Information Required In the Annotation of Models), defined by the European Bioinformatics Institute. In this context the databases CHEBI, GO, SBO and SGD are supported. The modules themselves are stored in relational databases: MySQL for concurrent access to the models by several user and Sqlite as the local travel database. In summary, the databases, MIRIAM registry and SBML feature allow scientists an interdisciplinary exchange of their models and make the comparison of their simulation results easy. For the simulation task one of currently five available solver plug-ins can be chosen: odeint, CVODE, COPASI and assimulo. The biological context for the mathematical models is given by the knowledge base. This is another MySQL database where all information about parameters and initial values are collected from publications. This way the user can describe the same model with different parameters, covering different trust levels. The merged output between knowledge base and model description becomes the final model.

Other features are:

• reaction balance analysis of metabolism
• parallel simulation multiple models
• real time consolidation of ODE models

A future goal for this meta platform is to simulate different types of models, like Boolean, stochastic models or FBA in combination with ODE models as well as separately.

The cell volume development is particularly important for the modeling of growing cells, since all concentration dependend reactions are affected by it. To get a better understanding of the volume development, it is necessary to combine different aspects, such as internal and external osmolyte concentration, cell wall mechanics or the water flux over an intact plasma membrane. If the cell geometry is assumed to be a simple sphere this spatial problem can be described with ordinary differential equations. In this project we build volume regulation models and test them against data we acquired in our own lab. These data are ranging from time courses of cell volumes monitored with bright field microscopy to mechanical cell wall properties obtained by AFM.

In order to better understand the physiological processes happening during one cell cycle and also due to the lack of appropriate, quantitative data in the literature, we are conducting elutriation experiments with a high time resolution over one entire cell cycle. From these we obtain a consistent dataset that quantifies the cellular proteome, transcriptome, and metabolome in cooperation with the Sauer and Dittmar groups. We also monitor physiological parameters of the cells such as volume, DNA content and budding state. In the future these data will be used to better parameterize the yeast whole cell model.

We develop a whole-cell model of protein translation in yeast. Our aim is to describe the dynamics of translation on the polysome and to gain an understanding of how efficiently the various resources (mRNA, ribosomes, tRNA) are utilized. This in turn leads to a range of interconnected questions, e.g. how is co-translational regulation influenced by the choice of synonymous codons, are some functional classes of proteins translated more efficiently than others, or how does protein translation vary across the cell development cycle. A discrete stochastic simulation method was used because in particular for low-abundance mRNAs (1...100 molecules) this formalism is likely to better capture the stochastic dynamics of the translation process than a formalism based on continuous concentrations, such as the ODE approach. The model has been fully developed and has been used to address biological questions. As it stands, it is also available for deployment in the context of a larger model, e.g. including transcription, or in the context of a general whole cell model.