Artificial Intelligence-Powered High-Content Analysis: Methodologies and Applications in Bioactive Compound Discovery from Natural Sources.

High-content screening (HCS) is a useful phenotypic drug discovery technology that combines automated microscopy, image analysis, and high-throughput experimentation to comprehensively characterize biological responses to diverse perturbations. This review summarizes the methodological fundamentals of high-content analysis, including image preprocessing, cell segmentation, feature processing, and downstream analysis, as well as the diverse phenotypic datasets generated from different biological models, perturbation strategies, and staining approaches. Recent advances in artificial intelligence, particularly deep learning, have improved cell segmentation, image representation learning, and phenotypic profiling, enabling more accurate and scalable analysis of HCS data. We further highlight emerging applications of AI-powered HCS in pharmaceutical research, with a particular focus on the discovery of bioactive compounds from natural sources. Finally, we discuss current challenges and future perspectives, including the construction of large-scale phenotypic databases, the integration of AI throughout the screening workflow, and the development of intelligent screening platforms. These advances are expected to accelerate phenotype-driven drug discovery and promote innovation in natural product research.