Molecular representations in AI-driven drug discovery: a review and practical guide

Area of Science:

• Computational chemistry and cheminformatics
• Drug discovery and development
• Artificial intelligence in medicine

Background:

• Technological advancements, including the computer revolution and high-throughput screening, enable computational analysis of bioactive molecules.
• The need for standardized, computer-readable molecular representations is critical for interdisciplinary scientific understanding.
• Numerous chemical representations have been developed due to computational progress and the complexity of molecular structures.

Purpose of the Study:

• To present popular electronic molecular and macromolecular representations used in drug discovery.
• To describe the applications of these representations in AI-driven drug discovery.
• To provide a guide on structural representations crucial for AI in drug discovery.

Main Methods:

• Review of popular electronic molecular and macromolecular representations.
• Focus on graph-based representations.
• Discussion of applications in AI-driven drug discovery.

Main Results:

• Identification of key molecular representations utilized in modern drug discovery.
• Demonstration of the utility of these representations in AI-driven drug discovery workflows.
• Highlighting the importance of graph-based representations.

Conclusions:

• Effective molecular representation is fundamental for computational drug discovery.
• Understanding these representations is essential for researchers applying AI in this field.
• This review serves as a foundational guide for novice researchers in computational drug discovery.