Sequential Design of Experiments enables accessible and efficient multidimensional mapping of single- and multi-protein condensates

Abstract

Biomolecular condensation emerges from networks of multivalent interactions that respond collectively to molecular composition and environmental conditions. Here, we establish a sequential Design of Experiments (DoE) framework to systematically and efficiently characterise multidimensional condensation landscapes. Applied to three RNA-binding condensate proteins – Dhh1, Ded1 and PGL-3 – the workflow revealed unexpected protein-specific interactions between biophysical and biochemical variables, including RNA, pH and ionic strength, and uncovered context-dependent switch-like behaviours. Extending the framework to protein mixtures revealed how proteins affected each other’s condensation and whether they formed condensate-within-condensate structures under different experimental conditions. Finally, an unexpected response identified by DoE prompted mechanistic analysis of Dhh1, revealing an ATP-independent RNA-binding surface that promotes efficient condensation in vitro and contributes to P-body formation in cells. Sequential DoE therefore provides an experimentally accessible strategy for uncovering context-dependent condensate regulation and converting multidimensional response patterns into mechanistic hypotheses.