The ability to accurately predict mechanism of action (MOA) of small molecules is essential for advancing drug discovery and repurposing efforts. To address this challenge, we aim to develop an AI-driven neural network framework that integrates chemical structures and morphological features derived from high-throughput fluorescent imaging data for MOA prediction. By leveraging computational resources and advanced neural network models, we strive to establish a robust predictive platform that can streamline the identification and repurposing of therapeutic agents.