A novel efficient drug repurposing framework through drug-disease association data integration using convolutional...

Area of Science:

• Computational Biology
• Bioinformatics
• Drug Discovery

Background:

• Drug repurposing offers a faster, cheaper alternative to de novo drug discovery.
• Traditional methods are costly, time-consuming, and have high failure rates.
• Data-driven approaches are emerging for identifying candidate drugs for specific diseases.

Purpose of the Study:

• To propose a deep neural network model, IDDI-DNN, for effective drug repurposing.
• To integrate diverse drug, disease, and association data for enhanced prediction.
• To identify novel drug-disease associations using advanced computational methods.

Main Methods:

• Constructed similarity matrices for drug properties, disease properties, and drug-disease associations.
• Integrated these matrices using a two-step similarity network fusion method.
• Employed a convolutional neural network for predicting unknown drug-disease associations.

Main Results:

• The IDDI-DNN model demonstrated high prediction accuracy.
• Comparative evaluations on two datasets (gold standard and DNdataset) were performed.
• IDDI-DNN outperformed existing state-of-the-art methods in predicting drug-disease associations.

Conclusions:

• IDDI-DNN is a powerful tool for drug repurposing.
• The model effectively leverages integrated data for accurate drug-disease association prediction.
• This approach accelerates the identification of potential new therapeutic uses for existing drugs.