Abstract
The success of artificial intelligence (AI) algorithms in predicting protein structure and more recently, protein interactions, demonstrates the power and potential of machine learning and AI for advancing and accelerating biomedical research. As cells are the fundamental unit of life, applying these tools to understand and predict cellular function represents the next great challenge. However, given the complexity of cellular structure and function, the diversity of cell types and the dynamic plasticity of cell states, the task will not be easy. To accomplish this challenge, AI models must scale and grow in sophistication, fueled by quantitative, multimodal data linking cell structure (their molecular composition, architecture, and morphology) to cell function (cell type and state). As cell biologists embrace the potential of AI models focused on cell features and functions, they are well positioned to contribute to their development, validate their utility, and perhaps, most importantly, play a leading role in leveraging the powers and insight emerging from the coming wave of cell-scale AI models.