Drug repurposing for non-small cell lung cancer by predicting drug response using pathway-level graph convolutional...

Area of Science:

• Computational biology
• Pharmacology
• Oncology

Background:

• Drug repurposing identifies new uses for existing medications.
• Predicting drug response is crucial for identifying viable candidates.
• Current deep learning models lack integration of biological pathway interactions.

Purpose of the Study:

• To develop an advanced drug response prediction model.
• To integrate biological pathway-pathway interactions into deep learning for enhanced prediction.
• To identify potential drug repurposing candidates for Non-Small Cell Lung Cancer (NSCLC).

Main Methods:

• Developed a novel deep learning model incorporating graph convolution operations.
• Utilized a pathway-pathway interaction network to represent cancer cell-line features.
• Applied the model to predict drug response on the GDSC1000 dataset.

Main Results:

• The proposed model demonstrated superior performance in drug response prediction compared to state-of-the-art methods.
• Achieved improved metrics including Root Mean Squared Error, Coefficient of Determination, and Pearson's Correlation Coefficient.
• Identified several potential drug repurposing candidates for NSCLC, with validation in clinical studies and literature.

Conclusions:

• Integrating pathway-pathway interactions significantly improves drug response prediction accuracy.
• The developed model is effective for identifying novel drug repurposing opportunities for NSCLC.
• This approach holds promise for accelerating drug discovery and development.