A Review Paper of Drug Disease Association Prediction
Yashika Sharma, Ronak Saraswat, Prince Raj, Sonia Lamba · 2025
The traditional drug R&D approach has great challenges in the past few decades, such as high cost, long development time, and high risk. Conventional design of pharmaceuticals is costly and time-consuming. To overcome these challenges, scientists and researchers are moving towards the computational approaches to predict relationship between drug-disease. Discovering new therapeutic applications for approved drugs not only accelerates treatment discovery but also significantly reduces the costs and time required for traditional drug development. During the COVID-19 pandemic, there was an urgent need for effective treatments, and remdesivir was successfully repurposed to combat the virus. In this paper, the previous computational methods for predicting drug–disease associations can be roughly divided into four categories, i.e. network based methods, machine learning-based methods, similarities-based methods and deep learning-based methods. This review examines different computational methods, highlighting their strengths, challenges, and the evolving trends in predictive modelling for drug development. This study explores advanced approaches for predicting drug-disease associations, with a particular emphasis on deep learning techniques.