The escalating global cancer burden has spurred extensive research and development efforts towards effective anti-cancer agents. Nevertheless, the exorbitant cost of developing novel drugs poses a significant challenge. This paper describes a cost-effective strategy, drug repositioning, which reuses approved drugs for novel medical indications, potentially resolving this predicament. This paper presents a comprehensive review of computational strategies for cancer drug repositioning, with an emphasis on machine learning. In recent years, the integration of bioinformatics technology and multi-omics data has significantly propelled the progress of cancer drug repurposing. In particular, machine learning and deep learning have contributed significantly to the remarkable advancements in cancer drug repositioning. This review summarizes the current application status of traditional computational methods and machine learning in this domain, and the analysis shows that machine learning holds great promise in facilitating cancer drug repositioning, both independently and in combination with other bioinformatics-based approaches. This paper can provide a valuable reference for the further integration of computational strategies and cancer drug research and development.

全球癌症负担的不断增加促使人们大力研究并试图开发更有效的抗肿瘤药物。然而，开发新药的成本昂贵。本文介绍了一种经济有效的策略——药物重定位，可以将已批准的药物重新用于新的适应证。本文对癌症药物重定位的计算策略进行了回顾，重点关注机器学习。近年来，生物信息学技术和多组学数据的融合极大促进了癌症药物的重新利用。机器学习和深度学习尤其促成了癌症药物重定位。本综述总结了传统计算方法和机器学习在该领域的应用现状，分析表明无论是独立的还是与其他基于生物信息学的方法相结合，机器学习在促进癌症药物重定位方面都具有很大前景。本文可以为癌症药物重定位的计算策略与癌症药物研发相结合的进一步发展提供参考。