Artificial intelligence-based decision support model for new drug development planning.

Area of Science:

• Pharmaceutical Sciences
• Computational Biology
• Machine Learning in Drug Discovery

Background:

• Drug development has a low success rate despite high potential returns.
• Existing strategies to improve success rates have been largely ineffective.
• Effective decision-making tools are needed for the early stages of drug development.

Purpose of the Study:

• To develop a machine learning model to guide effective decision-making in new drug development planning.
• To create a hybrid model for recommending and predicting suitable drug groups for pharmaceutical companies.
• To enhance the success rate of new drug development through data-driven recommendations.

Main Methods:

• Developed a hybrid model named Drug Development Recommendation (DDR).
• Integrated association rule learning, collaborative filtering, and content-based filtering.
• Utilized a random forest classification algorithm for content-based filtering, achieving 78% accuracy and 0.74 AUC.

Main Results:

• The DDR model demonstrated effectiveness in recommending drug groups for development.
• Applied to Coronavirus disease 2019 (COVID-19) vaccine development, higher DDR scores correlated with advanced clinical phases.
• The model supports rational decision-making by considering technical and company-specific variables.

Conclusions:

• The DDR model offers a novel approach to support strategic decision-making in early-stage drug development.
• It provides enterprise-customized recommendations by integrating multiple machine learning techniques.
• The model has the potential to improve the efficiency and success rate of pharmaceutical R&D.