ForSys: Non-invasive stress inference from time-lapse microscopy

During tissue development and regeneration, cells interpret and exert mechanical forces that are challenging to measure <i>in vivo</i>. Stress inference algorithms have thus emerged as powerful tools to estimate tissue stresses. Yet, effectively incorporating tissue dynamics into these algorithms remains elusive. Here, we introduce ForSys, a Python-based software that infers intercellular stresses and intracellular pressures from time-lapse microscopy. After validation, we applied ForSys to the migrating zebrafish lateral-line primordium, revealing increased stress during the cell rounding that precedes mitosis and accurately predicting the onset of epithelial rosettogenesis. We further used ForSys to study neuromast development and uncovered mechanical asymmetries linked to cell type-specific adhesion. The software performs both static and dynamic stress inference, supports command-line use, scripting, and a user-friendly graphical interface within Fiji, and accepts segmentation inputs from EPySeg and Cellpose. This versatility of ForSys enables the analysis of spatiotemporal patterns of mechanical forces during tissue morphogenesis <i>in vivo</i>.