eScience Center awards eight projects
The Netherlands eScience Center is pleased to announce eight winning project proposals. The awarded projects from the Open eScience Call (OEC) in 2024 will officially kick-off in 2025.
The annual eScience Center calls are aimed at researchers from Dutch research organizations, from any scientific discipline, that could benefit from the development and application of research software.
Each research team will be partnered up with eScience Center research software engineers (RSEs) who will advise on and co-develop sustainable research software, help organize community-oriented workshops, collaborate on research papers and much more. Learn more about how we collaborate on research projects here.
Learn more about the awarded projects
Emulating ice sheet models using graph networks (IceGraph)
Lead applicant: Dr. Stef Lhermitte (Delft University of Technology)
Discipline area: Natural Sciences and Engineering
Sea level rise projections are uncertain due to model limitations in simulating ice sheet mass loss from the Greenland and Antarctic Ice Sheets. While deep learning emulators offer a more efficient alternative, current models are unsuitable for ice sheets because they can’t be properly simulated on regular meshes. The IceGraph project, involving Delft University of Technology and partners, aims to develop a new emulator using MeshGraphNets to address this gap in ice sheet modeling.
TU Delft Astrodynamics Toolbox: bringing open science to spaceflight
Lead applicant: Dr. Dominic Dirkx (Delft University of Technology)
Discipline area: Natural Sciences and Engineering
Spaceflight dynamics requires high-fidelity software to process data on the motion of objects in space, but most existing tools are either closed-source or niche. This limits transparency, prevents full assessment of results, and hinders integration of diverse data sets and methodologies. We seek support from the eScience Center to professionalize the open-source TU Delft Astrodynamics Toolkit (Tudat), addressing current bottlenecks in software architecture and data processing to make the field more accessible, scalable, and open to the broader community.
From theory to gradients: crop growth models for the AI era
Lead applicant: Prof. dr. Ioannis Athanasiadis (Wageningen University & Research)
Discipline area: Environment and Sustainability
Together with the eScience Center, we will improve the crop growth simulation models using differentiable programming. We will develop a differentiable modeling framework and re-implement the well-known WOFOST crop growth model from Wageningen University. The framework will be demonstrated in two real-world use cases: sub-national yield forecasting for the European MARS Yield Forecasting System and AgMIP future climate impacts on yields. This tool will enable AI-driven advancements in crop modeling, combining deep learning with process-based components to address food security applications from breeding to precision agriculture and climate change.
Motion by Learning automaTion (MobyLeTion)
Lead applicant: Dr. Freek van Leijen (Delft University of Technology)
Discipline area: Environment and Sustainability
Satellite radar interferometry is a powerful tool for monitoring surface motion and infrastructure, but current methods lack systematic, informed monitoring. We are developing the CAROLINE software system to assess the dynamic behavior of structures using contextual information, addressing limitations in scalability and maintainability. By leveraging modern tools and GPUs, CAROLINE will enable efficient, continuous monitoring of critical infrastructure like dikes, roads, and bridges.
Biosynthetic genomics visualization and interoperability with enhanced workflows and scalable software
Lead applicant: Prof. dr. Marnix H. Medema (Wageningen University & Research)
Discipline area: Life Sciences
antiSMASH and BiG-SCAPE are user-friendly tools that help scientists worldwide to study microbial genomes, focusing on biosynthetic gene clusters (BGCs) that produce important natural molecules like antibiotics and crop protection agents. antiSMASH identifies BGCs in microbial genetic data, while BiG-SCAPE maps their diversity through network analysis. Our goal is to improve BiG-SCAPE’s scalability and interoperability with antiSMASH, and provide a combined user-friendly visualization tool that will allow an effective exploration, quantification and navigation of biosynthetic diversity across the biosphere.
Protein detective: a software for AI-driven identification of unknown proteins in cryo-EM density maps
Lead applicant: Prof. dr. Alexandre Bonvin (Utrecht University)
Discipline area: Life Sciences
Advances in Cryo-Electron Tomography (Cryo-ET) allow us to study biomolecules in their native environment, but identifying unassigned densities in intermediate-resolution maps remains a challenge. This project aims to develop a software pipeline that uses AlphaFold2 models, fast rigid-body fitting, and AI-based scoring to identify unknown proteins in Cryo-ET maps and build reliable structural models. The resulting tool will help the scientific community model protein complexes in medium-resolution Cryo-ET maps, enabling better identification of previously uncharacterized components.
Popylar: popularizing population receptive field modeling in neuroimaging
Lead applicant: Dr. Tomas Knapen (Vrije Universiteit Amsterdam & Netherlands Institute for Neuroscience )
Discipline area: Social Sciences and Humanities
Population receptive field (pRF) models are essential for understanding how the brain processes information across sensory domains, clinical conditions, and cognitive functions. The current pRF modeling software is outdated, leading to fragmentation and reproducibility issues. This project aims to create a unified, well-validated and open-source software package that will foster innovations in psychology and neuroscience. The new pRF modelling package will be supported by experts and the international community and extended with educational tools.
Dynamic3D: real-time simulation and analysis for 3D scholarly editions
Lead applicant: Dr. Costas Papadopoulos (Maastricht University)
Discipline area: Social Sciences and Humanities
While 3D visualization has greatly impacted humanities and social sciences research, current tools fail to capture the dynamic aspects of material culture, limiting the exploration of phenomena like light, sound, and movement. Dynamic3D aims to address this gap by integrating real-time simulation and analysis tools into 3D scholarly editions, enabling dynamic exploration of complex datasets. The resulting tools will be incorporated into PURE3D, a national infrastructure for 3D scholarship, offering innovative methods for engaging with cultural heritage.
Collaborate with us
The eScience Center helps researchers develop the open-source software they need. For an overview of the full breadth of our research software across all disciplines, check out our Research Software Directory.
Each year, the eScience Center publishes open calls for researchers in any scientific discipline. Interested in future calls? Follow us on social media (X, LinkedIn or Mastodon) or subscribe to our newsletter to be the first to know about a new call for proposals.
The project text has been shortened using ChatGPT. The AI-output has been verified for correctness, accuracy and completeness, where needed, and approved by the authors.