人工智能背景下生物信息学教学改革策略与路径构建

Construction of Teaching Reform Strategies and Paths for Bioinformatics in the Context of Artificial Intelligence

摘要: 新医科与新农科交叉融合背景下，生命科学数据呈爆发式增长，对生物信息学复合型人才的创新实践与数智素养提出了更高要求。作为生命科学类专业的核心交叉课程，生物信息学在传统教学中面临算法原理抽象深奥、代码跨度大导致学生接受困难，教学内容滞后于前沿技术、实践环节脱离真实科研等现实困境。以生成式人工智能与大语言模型为代表的智能技术，为破解上述困境提供了新范式。本文深入探讨了人工智能驱动生物信息学教学改革的必要性，并提出了构建人工智能辅助个性化自适应学习环境、设计智能体(Agent)协同的探究式实践、重构学习环境、实践模式、内容体系、评价机制及思政融合五位一体教学内容、建立人机协作的过程性多元评价体系、有机融入科技伦理与科学精神等教学改革策略。随后，以“基于大语言模型的基因家族分析”教学为例，详细阐述了人工智能驱动的教学实施路径与学习成果评估方案。本文提出的教学改革策略为生物信息学及相关交叉学科的教学创新提供了切实可行的实践指导。

Abstract: In the context of the interdisciplinary integration of emerging medicine and emerging agricultural sciences, the explosive growth of life science data has placed higher demands on the innovative practical capabilities and digital intelligence literacy of bioinformatics professionals. As a core interdisciplinary course for life science majors, bioinformatics education traditionally faces several challenges, including the abstract and sophisticated nature of algorithmic principles, the substantial programming requirements that create barriers to student learning, outdated teaching content that lags behind technological advances, and practical training that is disconnected from authentic scientific research. Generative artificial intelligence (AI) and large language models (LLMs) provide a new paradigm for addressing these challenges. This paper explores the necessity of AI-driven reform in bioinformatics education and proposes a teaching reform strategy, encompassing the construction of an AI-assisted personalized adaptive learning environment, the design of Agent-assisted inquiry-based practical learning, designing inquiry based practices for agent collaboration, reconstructing the learning environment, practical models, content systems, evaluation mechanisms, and integrating ideological and political education into a five in one teaching content, establishing a process oriented multi-dimensional evaluation system for human-machine collaboration, and organically integrating scientific ethics and scientific spirit. Subsequently, using “LLM-based gene family analysis” as a teaching case, the paper elaborates on the implementation pathway of AI-driven instruction and the corresponding learning-outcome assessment framework. The proposed teaching reform strategy provides practical and feasible guidance for educational innovation in bioinformatics and related interdisciplinary fields.