DeepPurpose: a deep learning library for drug-target interaction prediction.

Area of Science:

• Computational biology
• Drug discovery
• Bioinformatics

Background:

• Accurate prediction of drug-target interactions (DTI) is essential for accelerating drug discovery.
• Deep learning (DL) models show promise for DTI prediction but can be challenging for non-experts.
• A need exists for accessible and comprehensive DL tools in DTI research.

Purpose of the Study:

• To introduce DeepPurpose, an easy-to-use deep learning library for DTI prediction.
• To provide a flexible platform for building and training customized DTI prediction models.
• To demonstrate the state-of-the-art performance of DeepPurpose.

Main Methods:

• Development of a comprehensive DL library, DeepPurpose.
• Implementation of 15 compound and protein encoders.
• Integration of over 50 neural network architectures.
• Facilitation of customized DTI model training and evaluation.

Main Results:

• DeepPurpose demonstrates state-of-the-art performance on multiple benchmark DTI datasets.
• The library offers a wide range of customizable components for DTI modeling.
• Ease of use is a key feature for both computer scientists and bioinformaticians.

Conclusions:

• DeepPurpose provides a powerful yet accessible solution for DTI prediction.
• The library empowers researchers to develop and apply advanced DL models for drug discovery.
• DeepPurpose facilitates efficient and accurate DTI prediction, advancing the field.