Predictive machine learning models for rational permeability design in de novo macrocycle engineering: a review.

Area of Science:

• Drug discovery and medicinal chemistry
• Chemical biology
• Pharmacology

Background:

• Small molecule drugs have limitations in targeting intracellular proteins with complex interaction surfaces.
• Classical drug-like compounds struggle with large, shallow, or flexible protein targets.
• There is a growing need for alternative molecular classes to address previously undruggable targets.

Purpose of the Study:

• To highlight the potential of macrocyclic compounds in drug discovery.
• To discuss the challenges associated with macrocyclic drug design, particularly membrane permeability.
• To underscore the importance of macrocycles for targeting difficult intracellular proteins.

Main Methods:

• Exploration of chemical space beyond Lipinski's Rule of Five (bRo5).
• Analysis of macrocyclic compounds as an alternative molecular class.
• Identification of membrane permeability prediction as a key challenge.

Main Results:

• Macrocyclic compounds show promise for modulating difficult intracellular targets.
• These molecules expand the chemical space accessible for drug development.
• Accurate prediction of membrane permeability for macrocycles is a persistent limitation.

Conclusions:

• Macrocyclic compounds represent a significant advancement in targeting challenging proteins.
• Overcoming the challenge of membrane permeability prediction is crucial for the rational design of orally bioavailable macrocyclic drugs.
• Further research into macrocycle properties is essential for unlocking their full therapeutic potential.