Predictive identification of co-formers in co-amorphous systems

Area of Science:

• Pharmaceutical Sciences
• Materials Science
• Computational Chemistry

Background:

• Co-amorphous systems offer improved drug properties but their formation is complex.
• Predicting co-amorphous system formation is crucial for efficient drug development.

Purpose of the Study:

• To elucidate the properties of co-formers that favor co-amorphous material formation.
• To develop and validate a predictive model for co-amorphous system formation.

Main Methods:

• Partial Least Square - Discriminant Analysis (PLS-DA) was employed.
• 36 variables characterizing co-former and binding energy properties were analyzed.
• Model validation involved predicting co-amorphous formation with mebendazole and 29 co-formers.

Main Results:

• Key favorable properties include high average molecular weight and specific hydrogen bonding differences.
• Unfavorable properties include positive excess enthalpy of mixing/hydrogen bonding and large Hansen parameter differences.
• The developed model demonstrated a 90% predictive accuracy for co-amorphous systems.

Conclusions:

• Specific physicochemical properties of co-formers can reliably predict co-amorphous formation.
• The PLS-DA model provides a valuable tool for screening potential co-formers in drug development.