Artificial intelligence, machine learning, and drug repurposing in cancer

Area of Science:

• Computational drug discovery
• Pharmacology
• Biotechnology

Background:

• Drug repurposing offers a cost-effective strategy for identifying new therapeutic applications for existing medications.
• Machine learning (ML) and artificial intelligence (AI) are increasingly utilized for systematic drug repurposing, leveraging big data resources to accelerate and de-risk the drug development pipeline.

Purpose of the Study:

• To review supervised ML and AI methods for drug repurposing using publicly available data.
• To highlight the application of comprehensive target activity profiles for systematic drug repurposing.
• To discuss the applicability of these methods to various indications, including anticancer therapies and COVID-19 treatments.

Main Methods:

• Focus on supervised machine learning and artificial intelligence techniques.
• Utilizes publicly available databases and information resources.
• Employs comprehensive target activity profiles, including off-target information, to identify repurposing candidates.

Main Results:

• The reviewed methods are applicable to a wide range of indications, with a focus on anticancer drug therapies.
• Extending drug target profiles to include therapeutically relevant off-targets enhances the systematic repurposing process.
• Functional testing of patient cells provides valuable data for tissue-aware AI approaches in drug repurposing.

Conclusions:

• Supervised ML and AI methods, particularly those utilizing comprehensive target activity profiles, offer a powerful approach to drug repurposing.
• The integration of functional testing data can overcome limitations of genomics-only approaches in cancer drug repurposing.
• These computational strategies hold significant promise for accelerating drug discovery and development across diverse medical indications.