人工智能赋能药学本科生创新能力培养

Enhancing Innovation Capabilities of Pharmacy Undergraduates through Artificial Intelligence

摘要: 围绕药学本科教育中创新能力培养不足、实践资源受限与个性化指导薄弱等问题，本文在建构主义、情境学习理论和认知负荷理论的支撑下，构建人工智能赋能的药学本科生创新能力培养体系，并进一步聚焦基于氛围编程的分子设计实训核心模块进行剖析。该体系以AI支持的知识图谱、虚拟仿真平台与大语言模型辅助编程环境为基础，将个性化学习、真实任务建构、模型训练与科研表达贯通起来。重点模块以药物性质预测–候选分子筛选–结构活性关系解释为任务链，整合RDKit、DeepChem、Jupyter Notebook和LLM交互助手，引导学生经历数据清洗、分子特征提取、算法建模、结果可视化和迭代优化等环节。教学实践表明，该模式有助于降低药学本科生进入AI药物设计实践的技术门槛，增强其问题建构、数据思维、科研表达与跨学科协作能力，可为药学专业本科教育的智能化转型和创新人才培养提供可操作的路径参考。

Abstract: To address the insufficient cultivation of innovation capability, limited practical resources and weak personalized guidance in undergraduate pharmacy education, this paper constructs an artificial intelligence-empowered cultivation system supported by constructivism, situated learning theory and cognitive load theory, and further provides an in-depth analysis of the core module of ambient-coding-based molecular design training. The system connects personalized learning, authentic task construction, model training and research communication through AI-supported knowledge graphs, virtual simulation platforms and large language model-assisted programming environments. The core module follows the task chain of drug property prediction, candidate molecule screening and structure-activity relationship interpretation, integrating RDKit, DeepChem, Jupyter Notebook and LLM-based interactive assistance to guide students through data cleaning, molecular feature extraction, algorithmic modeling, result visualization and iterative optimization. Teaching practice indicates that this model can lower the technical threshold for pharmacy undergraduates to engage in AI-assisted drug design, while strengthening problem construction, data thinking, scientific communication and interdisciplinary collaboration. It provides an operable reference for the intelligent transformation of undergraduate pharmacy education and innovation-oriented talent cultivation.