Predictive biomarkers from functional analysis of static data

• Integration of static bulk RNAseq data with different sources of prior knowledge to derive a quantitative characterisation of different facets of the TME that is predictive of response to immune checkpoint inhibitors. Tools: EaSIeR, iHet. Papers: Lapuente-Santana et al, Patterns, 2021, Lapuente-Santana et al, iScience, 2024.
• Inference of personalised (ensembles of) cell-cell interaction networks from bulk RNA-seq data and prior knowledge using random graph models. Tool: RaCInG. Paper: van Santvoort et al, Cell Reports Methods, 2025.

Spatial multicellular models

• Development of computational approaches to quantitatively characterise spatial localization of cell types in the tumor microenvironment combining imaging modalities (radiomics images, pathology slides) and molecular data. Tool: SPoTLIghT. Paper: Lapuente-Santana et al, npj Precision Oncology, 2024.
• Development of agent-based models to study how interactions between different cells in the TME coordinate tumour development and response to treatment. Models: Prostate cancer development, Prostate cancer androgen deprivation therapy. Papers: Passier et al, Cancer Research Communications, 2023, van Genderen et al, npj Systems Biology and Applications, 2024. See our cover artwork in the MathOnco newsletter.
• Training of agent-based models for specific contexts using imaging data. Paper: Wang et al, Bioinformatics, 2026.

Diagnostic and prognostic biomarkers from liquid biopsies

• Development of machine learning models to use tumour markers (TMs) from liquid biopsies such as circulating tumour DNA (ctDNA) and proteins in the blood for diagnosis and monitoring of lung cancer. Model: Diagnostic lung cancer model. Paper: Visser et al, Lung cancer, 2023.

Mechanistic models from functional perturbation data

• Development of microfluidics technology for high-throughput perturbation screening of live cells from cancer patient biopsies. Papers: Eduati et al, Nature Communications, 2018, Yelleswarapu et al, JoVE, 2023 (including videos ).
• Inference predictive personalised mathematical models of intracellular signalling pathways based on logic ordinary differential equations to understand cellular mechanisms mediating differential drug response in patients. Tool and model: Active learning pipeline and CXCL9 model. Papers: Wang et al, Molecular Systems Biology, 2026, Eduati et al, Molecular Systems Biology, 2020.