Comparative analysis on artificial intelligence methods for DTI and DTBA prediction in drug repurposing

Abstract

Drug repurposing has evolved as an attractive approach in the search for new therapeutic applications that are shorter in development time and lower in cost. At the core of drug repurposing, the key challenge in this field is the accurate prediction of drug-target interactions (DTIs) and drug-target binding affinities (DTBAs). Various Artificial Intelligence (AI) techniques, including machine learning (ML) and deep learning (DL) methods, have proven to be significant in improving the prediction capability of the DTI and DTBA models. In this review, we provide critical insights into the current state-of-the-art AI methods used for the prediction of DTI and DTBA by highlighting major progress, bottlenecks, and potential future research directions. Classify these approaches according to their algorithmic framework, feature extraction methods, data source, and performance measures, and provide an extensive review of their strengths against limitations. Lastly, the limitations of current AI-assisted DTI and DTBA prediction methods in drug repurposing applications are summarized and highlight possible directions to address those challenges.