From computer-aided drug discovery to computer-driven drug discovery.

Area of Science:

• Drug Discovery
• Computational Chemistry
• Structural Biology
• Machine Learning

Background:

• Computational chemistry and structure-based design were historically supportive, not primary, drivers in small molecule drug discovery.
• Recent decades have seen significant progress in computational methods and structural biology techniques.

Purpose of the Study:

• To review the synergistic advances in computational chemistry, machine learning, and structural biology.
• To highlight their impact on accelerating small molecule therapeutic discovery.

Main Methods:

• Development of physics-based computational approaches for predicting drug properties (e.g., potency, solubility).
• Improvements in artificial intelligence (AI) and deep learning (DL) methods.
• Increased computational power via GPUs and cloud computing for in silico chemical space exploration.
• Advancements in structural biology, including cryogenic electron microscopy (cryo-EM) and protein structure prediction.

Main Results:

• Accurate prediction of diverse endpoints from potency to solubility using computational methods.
• Ability to explore and profile vast chemical spaces in silico.
• Access to numerous high-resolution 3D structures of drug-receptor complexes.
• Emergence of structurally-enabled computational methods as a key driver in drug discovery.

Conclusions:

• The convergence of computational chemistry, AI, and structural biology is transforming small molecule drug discovery.
• These integrated approaches are essential for hit identification, hit-to-lead, and lead optimization.
• Structurally-enabled computational methods are now a driving force for novel therapeutic development.