A network based clustering approach for in silico drug repurposing in Alzheimer's Disease

Alzheimer's is a terminal disease that affects the brain. Existing strategies for drug development to combat Alzheimer's Disease (AD) are time intensive and necessitate extensive on field trials. In this study, a network based clustering model is constructed using drug molecules of AD extracted from the CHEMBL database and analysed. Initially, the bioactivity data of Alzheimer's was obtained from the CHEMBL database and curated using Lipinski's Rule. Next, clustering was performed on this curated data using KMeans++ algorithm. The cluster interactions were then visualized and evaluated using network metrics in Gephi. The best drug molecules from each cluster were identified based on the percentage distribution of modularity class and selected to build the final strongly interacted network. Finally, from this network, the potent lead molecules are identified and proposed for treating AD.