**1.3 Overview of basis of drug repurposing**

Due to the immense financial costs and timescale associated with traditional drug discovery methods, it is natural to assume alternatives may be preferable, such as repurposing current therapeutics. As shown in **Figure 1**, repurposing known therapeutics follows a similar but truncated development cycle as traditional drug discovery. The timeline includes the discovery of a therapeutic target, usually a new therapeutic indication for a previously approved indication. Then, it is followed by clinical trials. Given that the information on the preclinical, pharmacokinetic, and pharmacodynamic are already known, the clinical trial phase moves faster. Following this step is filing for market approval as usual [18, 19]. This allows for a development time of 5–10 years compared to 12–15 years of development for traditional drug discovery (**Figure 1**). This also drives down the two major factors hampering novel drug development: cost of new therapeutics and time of discovery to market. Therefore, it is not surprising that an increasing number of therapeutics developed by the FDA or EMA are repurposed therapeutics [20]. Interestingly, the discovery of novel indications for therapeutics has been made through a multitude of approaches. Drug repurposing involves integrating data from multiple resources and the use of different approaches to allow for the discovery of novel indications. One major path for novel indications is model-based computation or *in silico* drug repurposing. This can include numerous screens of a therapeutic concerning its drug molecular targets, chemical structure, and signaling pathways to predict unknown targets or biomarkers for disease [18]. Besides computational modeling, high-throughput and/or high-content screening (HTS/HCS) of drug compounds are frequently used to screen known therapeutics for novel targets [21]. Additionally, *in silico* screening of known compounds can be used for molecular docking or binding-based studies [22]. Furthermore, *in silico* screening can also help with the computational prediction of novel metabolic pathways, signaling pathways,

and protein-protein interactions between diseases and known drugs [23]. Other major methods of screening for novel indications of current therapeutics include recently AI-based machine learning, which has been used to some effect to screen large groups of compounds for anti-Sars-Cov-2 inhibition [24]. Further predictive modeling of drug repurposing includes network modeling wherein networks of drugs, genes, and drug products, as well as their interactions and relationships can be modeled, allowing for a greater understanding of structure-guided targeting of therapeutics [25]. Furthermore, large-scale genome-based predictive modeling, such as genome-wide association studies (GWAS) can help predict potential novel therapeutic interactions [26]. Another aspect not commonly studied is the known side effects of therapeutics and how it can be used to identify novel therapeutic targets *via* computational modeling [27]. According to Sahragardjoonegani and colleagues, roughly two-thirds of new therapeutics approved by the FDA within the last 15 years have not been indicated for secondary indications besides their original purpose [28]. This creates a largely untapped field of current therapeutics that have not been studied in the context of other diseases. Overall, drug repurposing for therapy requires less time and cost for development and research than traditional drug discovery. Thus, it is an attractive approach for the discovery of new therapeutics.
