Automated Molecular Design in BRADSHAW, Applied to the Optimization of ERAP1 Inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Endoplasmic Reticulum Aminopeptidase 1 (ERAP1) is an emerging target for cancer immunotherapy and autoimmune diseases.
• Automated molecular design is increasingly utilized in drug discovery.

Purpose of the Study:

• To pilot the comprehensive use of automated molecular design on a medicinal chemistry project.
• To optimize inhibitors of ERAP1 using the BRADSHAW platform.

Main Methods:

• Utilized the BRADSHAW platform for automated molecular design.
• Conducted four iterations of in silico molecular generation, property prediction, and filtering.
• Performed multiparameter optimization of potency, physicochemical properties, and pharmacokinetics.

Main Results:

• Successfully refined Machine Learning (ML) models, enhancing scoring accuracy and compound quality.
• Identified in vitro and in vivo tool molecules targeting ERAP1.
• Achieved optimization of potency, physicochemical properties, and pharmacokinetics.

Conclusions:

• The BRADSHAW platform effectively enabled automated design for ERAP1 inhibitor optimization.
• Integration of automated design into medicinal chemistry projects is feasible and yields valuable tool compounds.
• Reflections on human factors and recommendations for future computational design projects were discussed.