Explainable Drug Repurposing Approach From Biased Random Walks.

Area of Science:

• Computational biology
• Pharmacology
• Bioinformatics

Background:

• Drug repurposing accelerates the discovery of new therapeutic applications for existing drugs.
• Current drug repurposing methodologies have limitations, necessitating innovative approaches.
• Integrating gene similarity and network-based algorithms can enhance drug discovery.

Purpose of the Study:

• To present a novel, efficient mathematical methodology for drug repurposing.
• To enhance the explainability of drug-gene-disease association recommendations.
• To evaluate the accuracy and computational efficiency of the proposed approach.

Main Methods:

• Utilizing gene similarity scores and biased random walks on drug-gene-disease association networks.
• Employing Markov chains to provide explainability for the random walk-based recommendations.
• Conducting performance evaluations and a rheumatoid arthritis case study.

Main Results:

• The methodology demonstrates accuracy in providing relevant drug repurposing recommendations.
• The approach is computationally efficient compared to existing state-of-the-art methods.
• Explainability of recommendations is achieved through the underlying Markov chain model.

Conclusions:

• The novel drug repurposing methodology offers an accurate and efficient solution.
• The approach provides explainable insights into suggested drug-gene-disease associations.
• This method holds promise for advancing drug discovery and development.