Patient data first
Produce large multimodal datasets from longitudinal patient samples and patient-derived model systems where evolution is measured over months and years.
The Virtual Evolving Cancer Cell Consortium
VEC³
Mission
Control cancer by modelling human somatic evolution with AI.
VEC³ is the first consortium that aims at creating AI models that describe how somatic cell populations and their microenvironment evolve in patients, connecting short-term treatment response with long-term emergence of drug resistance.
Predict resistance
Train AI models on both molecular response after perturbation and long-timescale evolutionary dynamics to predict when resistance will emerge and why.
Reveal vulnerabilities
Identify molecular vulnerabilities in resistant populations and support discovery research, drug target identification, and predictive biomarker studies.
Scientific vision
From patients to patient-derived model systems for cancer control.
Cancer treatment efficacy combines short-term response, such as killing cells with a drug, with long-term evolution, leading to the emergence of resistance. We aim at bringing both timescales into a single modelling framework.
Unlike virtual cell efforts built mainly on artificial perturbations in cell lines, we place patient data at the heart of the discovery effort and learn from evolving somatic cell populations in patient-derived systems.
The consortium develops high-throughput multimodal readouts, interpretable AI architectures, and a robust data platform for mechanistic discovery in oncology.
View consortium members