Dr Argyris Zardilis

I am a computational biologist interested in the scientific and methodological challenges of understanding multi-scale processes in development. My model system is early Arabidopsis thaliana flower development, which combines biological complexity with experimental tractability.

My work spans the molecular and mechanical signals that regulate growth and shape. On the molecular side, I led the development of a spatiotemporal atlas of early flower development integrating live imaging, cell tracking, and the expression of key developmental genes. On the mechanics side, I have also developed finite element mechanical models to show that mechanical stress patterns aspects of growth in early flowers, acting both through microtubule networks and by influencing developmental gene expression.

A growing thread in my work is the development of methods for inverse problems in morphogenesis: automatically extracting mechanistic biophysical models from large-scale multi-modal datasets combining 4D imaging and sequencing data. This includes learning unified representations across data modalities and differentiable physics-based inference to discover the mechano-chemical rules governing complex developmental processes. An example of this work was work presented at ICLR 2025.

During my PhD at the University of Edinburgh, I developed Chromar, a formal language for modelling multi-scale biological systems, and used it to build a multi-scale whole-plant lifecycle model of Arabidopsis extended to the population level.