This study benchmarks single-cell chromatin accessibility technologies for training deep learning models in regulatory genomics. Using large-scale datasets from mouse and Drosophila, it shows that increasing dataset size can offset lower per-cell coverage. The optimized HyDrop v2 protocol enables cost-efficient, high-quality data generation, supporting robust sequence-to-function modeling powered by high-performance computing.

Brain activity measured with EEG offers a non-invasive way to study cognitive function. This research uses machine learning and brain connectivity patterns to detect Alzheimer’s disease and frontotemporal dementia, two conditions that are difficult to distinguish clinically. The approach improves diagnostic insights while highlighting the role of brain networks in neurodegenerative disorders.

Magnetic nanoparticles with a flower-like shape are among the most promising candidates for cancer heat therapy — but why do they perform so well? Using large-scale GPU simulations on the VSC and MeluXina supercomputers, researchers from Ghent University and the University of Luxembourg pinpointed a precise size "sweet spot" of 100–150 nm where spin disorder dramatically boosts heating performance.

Researchers developed the Generalized Van Cittert-Zernike Schell propagator, a new algorithm for simulating partially coherent light with unprecedented efficiency. The method achieves up to 100× faster computation and 1000× higher accuracy compared to existing approaches, enabling advanced optical simulations for holography, imaging, and photonics using VSC’s high-performance computing infrastructure.

STREAM-PRS is a multi-tool pipeline that computes and compares polygenic risk scores using five widely used methods. It standardizes scores, applies population structure correction, and selects the best-performing model based on variance explained and predictive performance. The framework improves reproducibility and portability of PRS analyses across datasets and research settings.

Using large-scale molecular dynamics simulations, this research quantifies the self-diffusivity of boron in high-temperature CaO–SiO₂–Al₂O₃ slags relevant to silicon recycling. Enabled by VSC Tier-1 supercomputing resources, the study provides diffusion values consistent with diffusion-controlled experiments and clarifies the role of convection in industrial and laboratory systems.

Researchers used population pharmacokinetic modeling and Monte Carlo simulations to optimize fluconazole dosing in critically ill patients. By analyzing data from multiple studies, the work identifies body weight and continuous renal replacement therapy (CRRT) as key factors influencing drug exposure and proposes a model-informed dosing strategy to improve target attainment in intensive care settings.

Cancer cells evolve over time through genetic mutations that enable them to adapt, survive therapy, and become drug resistant. Understanding these evolutionary processes at high resolution is essential for developing more effective, targeted cancer treatments.

Schistosomiasis is a snail-borne parasitic disease affecting over 200 million people worldwide. This study examines how schistosome species, parasite development, and co-infection influence the microbiome of the freshwater snail Biomphalaria glabrata. Using 16S rRNA gene metabarcoding, it reveals species-specific microbiome disruption, developmental timing effects, and a stabilizing role of certain co-infections in host–parasite–microbiome interactions.

Metro 3, a proposed north–south automated metro line in Brussels, has faced major delays and cost increases, rising from €0.85 billion in 2009 to €5–7 billion today. This study provides the first reproducible, scenario-based accessibility analysis of the project. Results show modest average travel-time gains but significant benefits for lower-income districts, with the full line improving both speed and reliability across the city.