Vouchers IRP 2022
Vouchers IRP 2022

IRP vouchers awarded to 12 interdisciplinary research projects

Twelve research groups of the Faculty of Science have received a voucher of €50,000 each to work on an interdisciplinary research project in the field of Green Information Technology, Machine Learning in the Natural Sciences or Experimental Laboratories in Life Sciences. The project teams received the sum as part of the faculty voucher arrangement of the Interdisciplinary Research Platform (IRP). The IRP voucher will allow researchers of the Faculty of Science to start a pilot project with researchers of other research institutes of the faculty or with external collaborators.

Interdisciplinary collaboration

There were 23 submissions in the first call for vouchers of the Interdisciplinary Research Platform. These have been assessed on four criteria by an external selection committee consisting of Ewa Szymanska (Friesland Campina), Manfred Opper (TU Berlin, University of Birmingham) and Fred van Leeuwen (NKI, University of Amsterdam). The four criteria are: innovation, consortium partners, impact, and feasibility. As the overall quality of the submissions was very high, the faculty board decided to award twelve vouchers instead of six.


The IRP vouchers of 2022 were awarded to:

RNA Sequencing of the neuromuscular junction to evaluate the role of skeletal muscle gene expression in FUS-associated amyotrophic lateral sclerosis (ALS)

Amyotrophic lateral sclerosis (ALS) is the most frequent adult-onset motor neurodegenerative disease, resulting in paralysis and death due to respiratory failure. ALS is incurable and there is a clear unmet medical need, which is attributable to the incomplete understanding of the molecular pathogenesis of ALS. In order to develop future therapeutic approaches that could block the onset or the progression of ALS we must identify the molecular derailments leading to the pathology. In the project, we will investigate the role of skeletal muscle in the pathogenesis of ALS and by using cutting-edge subcellular RNA sequencing approaches, identify potential therapeutic targets in this tissue. The project will resolve an ongoing debate in the field: does skeletal muscle contribute to ALS pathogenesis?

Neuromorphic computational and data science: towards disruptively green computing

Modern computing systems consume large amounts of energy, reaching seven percent of the global electrical energy production today. Machine learning pipelines, especially the use of deep learning, are boosting the demand for computation at a pace that doubles every two to three months. Several applications in particle physics and computational condensed matter physics dramatically illustrate the need for a paradigm shift in hardware design that leads to a faster, more energy efficient and scalable computation technology. Neuromorphic computing hardware offers great potential for such a disruptive transition. With this project, we will demonstrate the feasibility of computing tasks that are practically impossible with non-neuromorphic hardware, by optimising specific physics use cases on standard hardware and comparing with the implementation on low-precision neuromorphic devices. We will also demonstrate the scaling for both hardware situations and identify the regime where neuromorphic hardware becomes disruptive.

  • Main applicant: Johan Mentink
  • Research institutes involved: IMM and IMAPP
  • External parties: SURF, IBM, University of Twente

Machine learning in natural science: bridging the gap between data and understanding

Machine learning is revolutionizing scientific research by greatly expanding our capacities to find signals and correlations in the massive amount of data nowadays produced in scientific observation and experimentation. An obvious next step is to employ AI in the construction of explanations that yield understanding of phenomena. However, this raises the fundamental questions of what the conditions for scientific understanding are, and whether AI can meet these conditions. In our research project philosophers, computer scientists and physicists join forces to address these questions, by investigating how scientific understanding is achieved in practice and how AI-driven science – especially physics – can contribute to achieving this aim.

VOCSENSE: Towards smart soil sensing to expedite the transition to a greener agriculture

With a growing population, soil health is increasingly important. However, past land use intensification, including fertilizer and pesticide use has significantly aggravated soil biotic diversity and functioning. Emissions of Volatile Organic Compounds (VOCs) from soil could be promising to monitor soil biological health. The unique profile of soil VOCs can be linked to soil microbial community composition and concurrent soil biological health. The current challenge is to link VOC profiles to soil health parameters that can be used by farmers. The VOCSENSE team will deliver step-change research on the fundamental relationships between soil biodiversity, VOCs and soil health parameters, to guide farmers in sustainable decision making.

Shining light on the dark matter of cell biology

The quantification of material properties of intracellular condensates, i.e. membrane-less organelles, and other intracellular structures is one of the biggest challenges in modern cell biology. Measurement of these properties is essential, as it is linked to various diseases. With this project we will trap and manipulate condensates in living cells with the goal to extract a comprehensive set of material properties. These properties include responses to active deformation, which is relevant for their behaviour in the active cellular milieu. The project will establish a pipeline that can be applied to many different systems. We envision future applications not only for other intracellular structures, but also tissues and organoids.

  • Main applicant: Jorine Eeftens
  • Research institutes involved: RIMLS, IMM

Essential oils as green pesticides for sustainable agriculture

The use of synthetic pesticides helps to maintain crop yields, but their use has detrimental effects on ecosystems and human health. Therefore, alternative plant protection strategies are urgently needed. Essential oils (EOs), i.e. aromatic, volatile liquids obtained from plant material, could be an excellent alternative to synthetic pesticides, as they are environmentally friendly, biodegradable and have a broad spectrum of activity against plant pathogens, including oomycetes and fungi. The effectiveness of EOs is mainly due to triggering resistance pathways within the host plant. Using downy mildew infection in grapevine as a case study, this project aims to understand how to induce plant innate immunity through EO application, use this knowledge to support environmentally friendly viticulture and develop EO-products for use in the field.

Towards a chemical self-organizing computer: Mathematical modeling of Marangoni flows

Sensing of bio-molecular input in diagnostics typically involves single-use devices, or complex instrumentation that relies on electronics. Novel computational mechanisms, featuring the interplay of chemical and hardware properties, open more energy- and feedstock-efficient pathways to process molecular information into material-based responses. In this project, we aim to mathematically model an autonomous system that can be trained to recognize different classes of complex, multidimensional chemical information and produce self-organizing structures on a two-dimensional substrate. The model will predict the Marangoni flow patterns amongst a two-dimensional array of sources and receivers, controlled by geometric and chemical external parameters. By using machine learning in this system, we will be able to map the emergent changes back to quantitative information about the input, demonstrating the principles of a sensing device.

  • Main applicant: Vanja Nikolić
  • Research institutes involved: IMAPP, IMM

Fast, continuous, and complete estimation of acoustic spectrotemporal sensitivity

Natural sounds consist of joint spectral and temporal modulations, and the human auditory system is tuned to such dynamically changing complex sounds. Conventional procedures to quantify hearing sensitivity ignore these sound modulations and instead focus on audibility of a small number of unmodulated, discrete frequencies. This focus prevents the diagnosis of speech-recognition impairment and central auditory processing disorder. For the patients suffering from these disorders – often children – it is essential to have an early and complete diagnosis to start treatment as soon as possible to improve the quality of life for the affected populations. To that end, we will develop a fast, objective method to estimate the sensitivity to spectrotemporal modulations of sounds by combining psychophysical and machine-learning methods in a novel fashion.

  • Main applicant: Snandan Sharma
  • Research institutes involved: DCN, iCIS

Expanding the molecular toolbox for the detection of active microorganisms in complex communities

Microorganisms are vital to sustain life on Earth and are found virtually everywhere. However, only a fraction of them is in a metabolically active state. Thus, besides ‘who is there?’ major questions for microbiology research are ‘who is metabolically active?’ and ‘what are they doing?’. As differentiating between active and inactive cells within a population is challenging, and linking function to identity in the absence of isolates even more so, this project will identify and develop promising clickable building blocks or substrate analogues to expand the molecular toolbox for (environmental) microbiologists. We will then apply these novel activity-based labelling techniques to understand the role of individual microorganisms in nitrogen removal from our water to safeguard our health and the environment.

  • Main applicant: Sebastian Lücker
  • Research institutes involved: RIBES, IMM

Hydrazine production from wastewater using the hydrazine synthase enzyme from anammox bacteria and a novel stomatocyte compartementalization strategy

Large scale anthropogenic input of nitrogen into the environment, mainly through fertilizers used in agriculture, has caused water eutrophication, contributed to global warming, and led to disturbance of the global nitrogen cycle. The EU has been implementing increasingly stringent directives to prevent the discharge of reactive nitrogen species to the environment. More stringent emission standards are also applied to wastewater treatment facilities. Anammox bacteria recycle the reactive nitrogen species ammonium back to harmless dinitrogen gas and can thereby remove fixed nitrogen from natural and manmade ecosystems and balance the disturbed nitrogen cycle. In the meantime, they produce hydrazine - a valuable commercial product used, among others, as blowing agent, antioxidant, and rocket fuel - as a free metabolic intermediate. By using the hydrazine synthase enzyme from anammox bacteria and encapsulating it with a novel stomatocyte compartmentalization strategy, this project could lead to a new biological hydrazine production green technology from wastewater nitrogen species.

Scent of killer: do malaria parasites produce and sense organic volatile compounds?

Despite decades of elimination efforts malaria remains one of the deadliest infectious diseases worldwide. Malaria is caused by unicellular, eukaryotic parasites, which have a complex life cycle. Communication between parasites and between the parasite and its hosts could be exploited both for intervention and diagnostic purposes, and very few molecules involved in such communication have been identified. This project explores the relevance of volatile compounds in the biology of malaria parasite. Do volatile compounds influence parasite growth, development or gene expression? Success of this project could offer exciting new opportunities for the development of breath-based diagnostic devices for malaria.

  • Main applicant: Richard Bartfai
  • Research institutes involved: RIMLS, IMM
  • External parties: TRopIQ, Radboud University Medical Centre

Predictive maintenance on the Dutch power grid using automated partial discharge detection

Maintenance of power grid is essential in preventing power outages, and is of major importance to both the society and the economy. Power grid operators use targeted measurement devices called smart cable guards to get real-time information on the power grid. Specific patterns of partial discharges precede failure in cable joints, and the smart cable guard is used to localise and prevent these failures. Interpreting such partial discharge patterns is a time-consuming operation. Simple mathematical models used for automated detection of partial discharge patterns are unreliable, frequently missing patterns or finding false positives. This project will demonstrate how convolutional neural networks for detection and localisation of partial discharge patterns can greatly improve upon the current detection method and establish a novel use-case for predictive maintenance applied to power grids.

Contact information