Application status of traditional computational methods and machine learning in cancer drug repositioning

The escalating global burden of cancer has spurred extensive research and development efforts aimed at discovering effective anti-cancer agents. However, the prohibitively high costs associated with developing novel drugs remain a formidable challenge. This paper describes a cost-effective approach, drug repositioning, which repurposes approved drugs for novel therapeutic indications, offering a promising solution to this dilemma. We present a comprehensive review of computational strategies employed in cancer drug repositioning, with a particular focus on machine learning. In recent years, the integration of bioinformatics technologies with multi-omics data has significantly advanced the field of cancer drug repurposing. In particular, machine learning and deep learning techniques have been instrumental in driving substantial progress in this area. This review summarizes the current application of traditional computational methods alongside machine learning in drug repositioning, highlighting the great potential of machine learning, both independently and in synergy with other bioinformatics-based approaches. The insights provided here offer valuable reference for further integration of computational strategies into the research and development of cancer therapies.