Structure, solubility, and permeability relationships in a diverse middle molecule library

Area of Science:

• Medicinal Chemistry
• Drug Discovery
• Computational Chemistry

Background:

• Middle molecules (500-2000 Da) present unique challenges in drug development due to their complex structures.
• Predicting druggability requires understanding their solubility and permeability properties.
• Existing methods may not adequately address the nuances of middle molecule characteristics.

Purpose of the Study:

• To establish a predictive methodology for the druggability of middle molecules.
• To investigate the relationships between structural features, solubility, and permeability in a diverse middle molecule library.
• To develop a computational approach for predicting passive diffusion in middle molecules.

Main Methods:

• Examined a library of 510 middle molecules (including natural product derivatives, non-natural products, peptides, depsipeptides, macrolides, and lignans).
• Assessed solubility and passive diffusion velocity using the parallel artificial membrane permeability assay (PAMPA).
• Correlated quantitative solubility and permeability data with molecular structural features.

Main Results:

• Obtained quantitative solubility data for 471 molecules and passive diffusion velocity for 287 molecules.
• Identified correlations between molecular structural attributes and measured solubility/permeability.
• Demonstrated the feasibility of predicting passive diffusion based on 3D structural features.

Conclusions:

• A novel method for predicting passive diffusion characteristics of middle molecules was developed.
• The methodology leverages structural features to forecast permeability, aiding in drug candidate selection.
• This approach enhances the potential druggability assessment of middle molecules in drug discovery pipelines.