Cell Physiology and Applications

Below, you can find the updated list of Bachelor and Master internships available at the moment.

For more information on available internships on Single-cell Physiology and on Host microbiome interactions, please contact directly the teamleaders in this group:

Jurgen Haanstra – j.r.haanstra[at]vu[dot]nl

Herwig Bachmann – h.bachmann@vu.nl

Remco Kort – r.kort@vu.nl

Project titleType of researchSupervisor(s)
Automated quantification and interpretation of GC-MS-based metabolomic data

between June 2021 and June 2022
Avis Nugroho (a.d.w.nugroho@vu.nl) & Herwig Bachmann
Volatile molecules are responsible for the sensory perception of a food product. In fermented foods, such volatiles can originate from the raw substrate itself, or from microbial activities. Over the course of fermentation, diverse enzymatic and chemical reactions occur, and the corresponding metabolites have various flavour characteristics and odour-activity thresholds. The analysis of these volatiles is done by GC-MS (gas chromatography - mass spectroscopy) which results in complex datasets. However, such analysis is mainly performed semi-automatically, requires laborious human curation, and consequently reports a small fraction of the detected compounds. Within this project, you will analyse a set of GC-MS data of fermented plant protein ingredients and focus on the untargeted analysis of the data with the aim of identifying the entire remaining compounds that are currently not reported. The project requires affinity with complex data analysis and independent learning to work with different software packages. Furthermore a basic understanding of analytical chemistry - in particular GC-MS - would be beneficial.

Task summary:
Within this project, you will explore and compare various approaches to automate characterization and quantification of GC-MS peaks, with and without a dedicated software. For the dedicated software, you will receive a training from an expert at the beginning of your project. This will ideally result in a nearly fully automated data analysis pipeline starting from the raw GC/MS data and ending in a comprehensive overview of detected compounds. If time allows, we will explore the integration of the metabolomics (volatolomics) data with their corresponding genome data. Next to our supervision, you will be working in close collaboration with GC-MS experts from NIZO Food Research.
If you are interested or would like to discuss about this project, do not hesitate to contact Avis Nugroho (a.d.w.nugroho@vu.nl).
Production of cultured red blood cells for transfusion purposes: analysis of metabolomics data to achieve high cell density erythroblast cultures Theoretical (Master)Jurgen Haanstra (in collaboration with Sanquin)
Transfusion of donor-derived red blood cells (RBC) to alleviate anemia is the most common form of cellular therapy. In addition, red blood cells hold great promise as delivery agents of e.g. specific drugs or enzymes. However, the availability of transfusion units depends on volunteers and carries a potential risk of alloimmunization and blood borne diseases. More than 30 bloodgroup systems encode >300 bloodgroup antigens and bloodgroup matching becomes increasingly challenging in a multiethnic society. Particularly the chronically transfused patients are at risk for alloimmunisation. In vitro cultured, customizable red blood cells (cRBC) would negate these concerns and introduce precision medicine both in transfusion medicine as well as in drug delivery applications. We aim to produce human cRBC at large-scale and cost effective, for which we need to optimize culture conditions and reduce cost-drivers.
Transfusion-ready erythrocytes can be cultured from hematopoietic progenitors but at market-incompatible high costs. A limitation in maximum cell density, 2 million cells/mL, has been observed in in vitro erythroblast expansion. Understanding the origin of this cell density limitation may provide strategies, both at media composition and feeding regime levels, to facilitate economically feasible upscaling.
Analysis of cell-conditioned media indicated that small molecules (<3kDa) are responsible for growth limitation. A metabolic by-product may be the culprit. Alternatively, or in addition, depletion of nutrients may also contribute to the growth stop. Therefore we aim to analyse the metabolic activity of erythroblasts with the aim to adapt the media such that cells can be cultured at much higher densities.
We have produced transcriptome and proteome data from which we can deduce the metabolic pathways that are active in our erythroblast cultures. We also determined metabolic profiles of erythroblasts seeded in defined medium, and the corresponding profiles of the medium, and sampled at timepoints 0, 12, 24 and 36 hours.
In this project you are going to use the genome-scale reconstruction of human metabolism. The transcriptome and proteome data will be used to restrict the enzymes in this model to what is actually expressed in erythroblasts. The measured metabolic profiles will be either used as input to understand internal flux distributions and elucidate potential unwanted byproducts - or to compare them to predictions of flux distributions that would give optimal growth. Furthermore, investigating theoretical flux profiles that would give high growth rates (i.e. biomass production) will reveal options to adjust the culture media.
We are looking for an enthusiastic Master student with a bioinformatics background and an interest in metabolic networks

This project will be conducted in close cooperation between Sanquin Research, dept Hematopoiesis, and the Amsterdam Institute of Molecular and Life Sciences (AIMMS), Systems Biology Lab.
Integration of quantitative multi-omics data into genome-scale metabolic models Theoretical (Master)Pranas Grigaitis
Eunice van Pelt-KleinJan (e.van.pelt-kleinjan@vu.nl)
Computational models of microbial metabolism are useful tools in biotechnology and medical sciences due to their ability to predict and analyze microbial cell behavior in silico. Stoichiometric modeling is an attractive technique to use knowledge-bases to aid analysis of microbial metabolism at genome-scale level. However, these models have limited predictive power in a number of situations due to the assumption of analyzing (1) optimally-functioning networks in (2) a steady state, fully driven by reaction stoichiometry. Recent approaches to improve the predictions of these models usually rely on the detailed descriptions of protein turnover costs (proteome-constrained, pc-Models), and would provide a platform to aid the model by using quantitative -omics data.

In this project, we want to develop a framework of straightforward and standardized integration of multiple types of -omics data, namely, transcriptomics, proteomics and fluxomics.

Planned activities (and methods)
- Automatizing the integration of RNA-seq and mass spectrometry-driven label-free quantitative proteomics experiments into pcModels of Saccharomyces cerevisiae, Lactococcus lactis and/or Schizosaccharomyces pombe (scripting: Python, bash)
- Simulation of cross-condition pcModels on both local machine and compute cluster (Lisa/SURFsara) and biological interpretation of the resulting simulation results (scripting: Python, bash; data analysis: Python, R)

Three months