Identifying drug-target interactions based on graph convolutional network and deep neural network.

Area of Science:

• Computational chemistry
• Bioinformatics
• Drug discovery

Background:

• Identifying new drug-target interactions (DTIs) is crucial but challenging in drug discovery.
• Existing computational methods often model drug and target networks separately, neglecting drug-protein pair (DPP) associations.
• There is a need for novel approaches that integrate DPP associations for improved DTI prediction.

Purpose of the Study:

• To propose a novel computational framework, GCN-DTI, for identifying drug-target interactions.
• To incorporate associations between drug-protein pairs (DPPs) into DTI prediction models.
• To enhance the accuracy and efficiency of drug discovery pipelines.

Main Methods:

• Constructed a novel network where nodes represent drug-protein pairs (DPPs) and edges represent their associations.
• Developed a graph convolutional network (GCN) to learn feature representations for each DPP.
• Utilized a deep neural network (DNN) that takes DPP feature representations as input for final DTI prediction.

Main Results:

• The GCN-DTI framework demonstrated superior performance compared to existing state-of-the-art methods.
• The integration of DPP associations significantly improved the accuracy of DTI identification.
• The proposed model effectively learns complex relationships within the DPP network.

Conclusions:

• GCN-DTI offers a powerful and accurate computational approach for identifying novel drug-target interactions.
• This method addresses limitations of previous DTI prediction strategies by incorporating DPP associations.
• The framework has the potential to accelerate the drug discovery process by reducing time and cost.