top of page

We are looking for 15 Early Stage Researchers to join our project at multiple sites in EU with a master degree in a relevant discipline (biology/bioinformatics/mathematical/computational sciences) interested to develop innovative modelling techniques aiming to reconstruct the dynamics of complex multiscale biological systems (All the information is available on this website). ESR1, ESR2, ESR3, ESR4, ESR5, ESR6, ESR7, ESR8, ESR9, ESR10, ESR11, ESR12, ESR13, ESR14, ESR15.

There will be 2 rounds of applications

1st Round:


15th March-15th April (Applications) and 15-30 April (Selection)

  • ESR13 - CLOSED

February (Applications) and 1-15 March (Selection)

  • ESR5, ESR10, ESR15 - CLOSED

February, March (Applications) and 1- 15 May (Selection), the 

2nd Round:

1st May - 31st May (ESR4, ESR12: Applications) and 1-15 June (Selection) 

Selection Process has been prolonged for:

1st May - 15th June (ESR 2: Applications) and 15-30 June (Selection)

1st May - 23th June (ESR 3: Applications) and 23-30 June (Selection) 

1st May - 30th June (ESR6, ESR7: Applications) and 1-31st July (Selection)

1st May - 18th July (ESR8 and ESR9: Applications) and rolling basis (Selection)

28th September - 15 October (ESR11 and ESR14) and 15 - 31 October (Selection)

  1. Candidates apply for a position using the online application form at E-MUSE website               

  2. The E-MUSE Project Manager provides a first screen of the written applications to check eligibility of the candidate and forwards the eligible applications to the ESR supervisors. For ESR1, ESR4, ESR11, ESR12, ESR14: link for application on VUA, CHRH, UM and KUL recruitment platform respectively.

  3. The Supervisors per ESR will select the best candidates based on CV, academic records, recommendation and motivation letters and adequate skills set. To better assess the best candidate the shortlisted candidates might be asked to write an abstract of provided scientific documents relevant to the research subject.

  4. The selected applicants will be interviewed through an online meeting by Selection Committee per ESR.

  5. Optional: If possible, personal interviews will be held with the presentation of a master thesis by shortlisted candidates. Expenses made by the candidate for travelling and the hotel will be reimbursed by the local organisation. Due to Covid-19, this stage can be done via an online meeting.

  6. The best candidates will be chosen by the ESR supervisors and the Selection Committee(s). European Project Manager will communicate the successful candidates to the Consortium and Partners.

Recruitment Process

Job Interview

Multi-omics data mining through (dynamic) mechanistic models

Integration of omics data with genome-scale models to microbial ecosystems dynamics. The task involves the definition and solution of constraint-based models and objective functions integrated over dynamic periods. 

Dynamics of cellular interactions

The objective is to develop a dynamic model of microbial metabolism of cellular populations by using omics data. The following sub-objectives are to be accomplished: (1) the modelling of population growth taking into account competition/cooperation mechanisms between cells; (2) the modelling at the metabolic level by means of adequate constraint modelling based approaches; (3) to develop a multi-scale model to describe the microbial consortium dynamics in cheese ripening.

Feed-back control of individual cells in microorganism populations

Project aims at developing a novel control paradigm to guide the modulation of the underlying stochastic gene circuits of bacterial systems that regulate metabolic activity and therefore phenotype heterogeneity, through the available environmental stress variables. Central questions to be addressed concern, on the one hand, the appropriate system representation for control, which also includes uncertainty compensation for robust tracking and regulation. On the other hand, the availability of technologies and inference methods for heterogeneity quantification and monitoring. The proposed control paradigm will be demonstrated on cheese ecosystems, controlling environmental variables towards specific final product characteristics.

Systems-level physiological characterization of Lactobacillus helveticus

(1) Detailed physiological characterization of selected strain(s) of Lactobacillus helveticus, which is an industrially important lactic acid bacterium commonly used in the ripening of different types of cheese; (2) Identifying the biochemical basis and regulatory mechanisms underlying the organism’s contribution to cheese flavour; (3) Simulating growth and flavour compound production under industrially relevant conditions through constraint-based modelling.

Metabolomics and kinetic studies of microorganisms involved in cheese ripening

The objectives of the ESR project are (1) to study the effect of iron on growth and metabolism of ripening microorganisms, a predictive growth model will be done to better anticipate complex ecosystem understanding; (2) to analyse the metabolome of key microorganisms alone, and in consortia in controlled condition; (3) to constrain a genome-scale metabolic model in collaboration with ESR1 (VUA).

Development of kernel approaches for the integration of biological data from heterogeneous sources

Advances in biological data generation, such as next generation sequencing techniques and mass spectrometry, have made it possible to obtain a wealth of high dimensional, diverse biological information, including an individual’s genome, transcriptome, proteome, and metabolome. In this context, the ESR6 project will investigate non-linear multi-omics data integration with kernel approaches. Even if methodologies and tools exist for the integration of several datasets in a nonlinear way, a lot of work is required to make these methods really operational and efficient. Furthermore, methodological developments to address sparse approaches in kernel-based methods are still to be improved. A thorough study of ad hoc kernels for time course experiment and compositional data would also benefit to the scientific community. This project aim at bridging the gap between methodological developments and relevant application in systems biology. We will show how to take benefit from big amounts of data that can be generated and/or acquired in a systems biology approach in research.

Network-based approaches for multi-omics data analysis

To learn, implement and apply network-based algorithms to available data, embedding multi-omics data into a network framework. The ESR will deal with omics experimental data and combine them with relevant information about bacterial and metagenomics networks (e.g. protein interaction, biochemical pathways, metabolic networks). Starting from these reconstructed networks, several methods will be applied to identify key players (nodes, pathways, communities) related to the experimental conditions studied.

Machine learning methods at the genome scale

The objective of this ESR project is developing and fine tuning advanced machine learning models using multi-omics data. Development of mechanistic models will also be supported by machine learning, for instance by pinpointing inconsistencies in the models with the data and by identifying links between variables not captured in the models. Traditional machine learning approaches include unsupervised learning and clustering to reason about the network links and role in genome-scale models. Tree based methods (random forest, decision tree) provide a natural way to analyse the contribution of different features in a model. More recent sequence-based learning methods like recurrent neural networks are able to use contextual information. In these models there is no need for advance feature selection, since it is encoded in the weights of the network. Although the interpretability of these models is limited, in practice they perform well in case of large data available. Word/sentence embedding methods are state of the art ways to represent sequential data to deep neural networks in analogue natural language processing tasks. In this project sequence embedding techniques are investigated as an input layer of neural networks. The machine learning models are utilized in dynamics modelling and process modelling tasks.

Deep learning model development to predict cheese properties

The main objective of the ESR project is developing deep learning models using multi-omics data to estimate product properties at higher levels (population and/or macroscopic level). At the same time models will be trained to provide a priori knowledge for network-based approaches. For this purpose, ESR will explore novel neural network architectures and learning approaches. This includes experiments with different model architectures such as recurrent, convolutional (e.g. for the determination of protein-protein interactions), attention-based networks. Graph neural networks are recently proposed in systems biology, where a weighted graph is a commonly used representation. Ensemble learning embraces the decision of several machine learning models. We plan to experiment with the recent idea of ensemble of ensembles combined with unsupervised methods. Multi-task learning is a promising way to improve model performance where different learning tasks need similar input data, since this improves the generalization capability of models. This fits to the case of different biomarkers identification based on multi-omics data.

Cheese microbial ecosystem structuration and resilience

The objectives of the ESR project are to understand the effect of iron on the cheese ecosystem. Dynamic data will be collected during the ripening process of a model cheese produced with/without iron supplementation. Several macro-scale data will be collected: microorganisms’ growth, substrate degradation, pH, colour, rheology, moisture, organoleptic properties, and online measure of respiration. The deduced macroscopic properties will be used by the modelling researchers to identify relationships with micro-scale features from multi-omics data collected in parallel (e.g. transcriptome, metabolome). The ESR will focus his/her studies on biochemical pathways involved in iron homeostasis, oxidative stress, sulphur metabolism and volatile sulphur compounds production to understand the effect of iron on the ecosystem’s equilibrium. These metabolisms are interconnected and largely contribute to important organoleptic properties of fermented foods through the release of odor-active compounds.

Biology driven complexity reduction of individual based models

ESR11 will focus on the power and limitations of two predictive microbial modelling paradigms: on the one hand Individual-based Models (IbM) and on the other hand Partial Differential Equations (PDEs). In the context of multi-organism populations, Individual-based Model (IbM) approaches can cope with the spatial distribution that will be useful to describe cell to cell interaction. A typical challenge is to provide a bridge between the (computationally very intensive) IbMs and population models which typically take the form of PDEs, and therefore to provide a feedback to genome scale modelling with the help of accurate spatial/temporal characterizations.

Development of multi-scaling strategies

The main objective will be to analyse a cheese fungal/bacterial ecosystem and understand their functioning and impact on the ripening of fermented food products. The experimental data will be analysed with models describing either the kinetics, probability of growth/no growth in function of time or preselected conditions. This information will be integrated, through correlation of omics indexes with the population dynamics.

Fermentation and flavour development of plant-based cheese analogues using different starter cultures

The aim of this project is to predict key properties of starter culture strains for production of plant proteins cheese analogues with optimal fermentation profiles and flavour development. For approximately 100 lactic acid bacteria with sequenced genomes, phenotypic data will be generated during fermentation of different plant based substrates using high-throughput screening methodologies. Key phenotypic data include acidification rates, proteolytic capacity and development and presence of (precursors of) major flavour compounds. The impact of the fermentation process will also be assessed with respect to potential spoilage issues by sporeformers and safety issues by Listeria monocytogenes as influenced by production of acids and potential other inhibitory compounds. Random forest analyses will be performed using genotypic strain data, phenotypic data and fermentation conditions in relation to favourable fermentation outcomes (acidification, desirable flavours) and unfavourable outcomes (spoilage/safety). Using machine learning, the ultimate aim is to predict strain- and condition-dependent properties to produce different plant based cheeses.

Modelling the effects of ripening conditions on flavour development of semi-hard cheese and plant-based cheese analogues

The objective of this work is to generate a predictive model for the impact of ripening conditions on flavour development of semi hard cheese and plant based cheese analogues (or hybrids thereof). The building of the model will start with semi-hard Gouda cheese, with a focus on ripening parameters (e.g. temperature, time, aw) and the effects of these parameters on determined flavour development (sensory and chemical data). This will be done in close collaboration with NIZO, based on available data (literature) and expert knowledge present at NIZO. To extend application of the ripening model with variables that are important for new types of plant-based cheeses, additional experimental data will be generated in vitro with high through put methodologies. Using selected starter cultures for production of laboratory scale plant-based cheese (interaction with ESR13), the effect of various ripening conditions on flavour development and microbiological stability and safety will be assessed. To generate the in silico model, correlations between ripening conditions, cheese properties and the final flavour and texture features of the product will be investigated using linear (e.g., PCA, PLS) and non-linear (e.g., mutual information) methods. Based on the identified relations and the data availability, a regression model based on a machine learning method (e.g., Artificial Neural Networks, Random Forest, Support Vector Machine) will be trained and validated for the prediction of the final properties of the cheese analogues.

Dynamic modelling of use properties in cheese ripening

The objective is to develop dynamic models for the evolution of selected use properties of cheese during the ripening process. Use properties include sensory evaluations of cheese attributes expected by consumers, and technological properties desired by the cheese-making companies. Relations with microbiological, physical and chemical models developed in other ESRs’ projects will be established.

bottom of page