van der Weide Lab at UMC Utrecht

Chromatin Systems Biology

Connecting the dots between chromatin form and function

Every cell in the body carries the same genome, yet each one reads it differently. Which genes are used, and when, depends on many regulatory layers working together: how DNA is packed, which marks decorate it, where it sits in the nucleus, and what gets transcribed. When these layers fall out of sync, cells can lose their identity, stop dividing, or grow out of control. Yet we usually measure them one at a time, in different cells, and try to piece the picture together afterwards.

At the van der Weide Lab, we measure these layers together in the same single cell, and use computational biology and machine learning to understand how they work as a system. We focus on cells whose chromatin breaks the usual rules, because these exceptions show what each layer really does, and how its disruption contributes to disease.

Unusual biology often needs unusual tools, so we develop both.

Research

Our research asks how the many layers of gene regulation work together in a single cell, and what happens when they don’t. We want to understand gene regulation as a system, and use that to explain what goes wrong in disease. Our work rests on three connected pillars: single-cell methods that measure these layers together, computational and machine-learning approaches to connect them, and cells with non-canonical chromatin, where the usual rules break down.

1. Reading many layers of information in a single cell

Illustration of MAbID: antibody-DNA conjugates labeling DNA nucleosomes, alongside a molecular structure visualization of an antibody-DNA complex

Gene regulation runs on many layers at once: chromatin marks, DNA-binding proteins, nuclear organisation and transcription. Most methods measure one layer at a time, and often in bulk, so it's hard to see how the layers combine or how they differ between cells.

We use and develop single-cell methods that profile dozens of DNA-protein interactions together with the transcriptome in the same cell, across many samples in a single experiment. This lets us ask questions that were out of reach with one-mark-at-a-time approaches.


📄 Lochs, van der Weide et al., Nature Methods, 2024 

2. From many layers to one regulatory system

SCIBED workflow diagram showing dataset simulation, imputation of missing data, and downstream effects on ground-truth similarity, signal enrichment, and cell-subpopulation separation.

Measuring many layers in the same cell only pays off if we can connect them. Sparse signal, multiple modalities and technical artefacts make that hard, and make it easy to miss real biology or to find biology that isn't there.

We build computational and machine-learning approaches to integrate these layers and model how they work together, from imputing sparse single-cell data to learning which regulatory features shape gene expression. We also work as the computational partner for labs launching new assays.


📄 Moreno-González, […], van der Weide, NARgb, 2025 

3. Learning from cells that break the rules

Diagram of the vertebrate retina showing rod and cone photoreceptors, retinal neurons, and conventional versus inverted nuclear organization.

In a textbook nucleus, active DNA sits in the centre and silent DNA is packed against the edge. Rod photoreceptors in nocturnal animals flip this around. Senescent cells fold their chromatin into dense foci. These systems are often treated as curiosities.

We see them as natural experiments. By applying our single-cell and computational tools to them, we ask which chromatin features really drive gene regulation, and what happens when they fail.

Meet the Team

Publications

  • A computational framework to dissect imputation strategies for single-cell histone modification data

    Marta Moreno-González, Jeroen de Ridder, Jop Kind, Robin H van der Weide

    NAR Genomics and Bioinformatics, December 2025. PMID: 41472883

    PDF
  • Combinatorial single-cell profiling of major chromatin types with MAbID

    Silke J. A. Lochs*, Robin H. van der Weide*, Kim L. de Luca, Tessy Korthout, Ramada E. van Beek, Hiroshi Kimura & Jop Kind

    Nature Methods, January 2024. PMID: 38049699

    PDF

News