Statistical analysis and compound selection of combinatorial libraries for soluble epoxide hydrolase

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Soluble epoxide hydrolase (sEH) inhibitors are pursued for treating hypertension and cardiovascular/renal diseases.
• Combinatorial chemistry and QSAR modeling are key in drug discovery.

Purpose of the Study:

• To develop robust QSAR models for diverse sEH inhibitors.
• To identify novel, high-activity sEH inhibitors using computational approaches.

Main Methods:

• Generated QSAR models using Daylight, MOE 2D, and DragonX descriptors.
• Trained and validated Gradient Boosting Machines (GBM), Partial Least Squares (PLS), and Cubist models.
• Employed a consensus approach and Gaussian process modified sequential elimination (G-SELC) for chemical space expansion.

Main Results:

• Achieved high Q(2) and R(2) values for GBM, PLS, and Cubist models.
• The consensus model demonstrated robust prediction on an external validation set.
• Identified 50 new, structurally diverse compounds with predicted high sEH inhibitory activity using G-SELC.

Conclusions:

• The developed consensus QSAR model effectively predicts sEH inhibitor activity.
• G-SELC is a valuable method for expanding chemical space and designing targeted compound libraries.
• This balanced approach accelerates the identification of promising drug candidates for laboratory synthesis.