Artificial intelligence for multiscale mechanobiology

Mechanobiology investigates how mechanical forces influence cellular behavior across scales, from molecular interactions to tissue-level responses. Advances in force measurement, high-resolution imaging, and multiomic profiling have generated datasets of increasing dimensionality, modality, and scale. Artificial intelligence (AI) and machine learning (ML) methods are well suited to analyzing such data, predicting cellular responses to mechanical stimuli, and identifying candidate mechanistic relationships. Here we review key applications of AI and ML in mechanobiology, including force inference from imaging data, tissue stiffness mapping, cell phenotype classification, chromatin-based prediction of protein localization, and computational modeling of mechanosensitive protein dynamics. We discuss emerging tools such as foundation models for cell segmentation and protein structure prediction, as well as the integration of AI-predicted mechanical properties with spatial transcriptomics. We also examine current limitations—including data scarcity, the gap between proof-of-concept and validated tools, and the challenge of causal interpretation—and identify opportunities for real-time closed-loop systems, therapeutic translation, and multiscale modeling.