Measuring experimental cyclohexane-water distribution coefficients for the SAMPL5 challenge

Area of Science:

• Physical Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Distribution coefficients quantify small molecule preference between immiscible phases (e.g., cyclohexane/water).
• These coefficients test thermodynamic and chemical accuracy of physical models, crucial for drug discovery.
• Traditional octanol-water systems present challenges due to high water content and heterogeneity.

Purpose of the Study:

• To assess distribution coefficients as a method for evaluating and improving predictive physical models.
• To conduct a blind prediction exercise (SAMPL5 challenge) using cyclohexane/water distribution coefficients.
• To establish a reliable experimental protocol and data analysis for cyclohexane/water distribution coefficients.

Main Methods:

• Utilized a modified shake-flask liquid chromatography-tandem mass spectrometry (LC-MS/MS) protocol.
• Measured cyclohexane/water distribution coefficients for 53 druglike compounds at pH 7.4.
• Developed a bootstrap-based data analysis to account for experimental errors.

Main Results:

• Generated experimental cyclohexane/water distribution coefficient data for 53 compounds.
• 18 research groups participated in the blind prediction challenge prior to data release.
• The study provides a robust dataset and methodology for model validation.

Conclusions:

• Cyclohexane/water distribution coefficients offer a valuable route for evaluating predictive physical models.
• The experimental data and analysis methods will guide future improvements in computational chemistry for drug discovery.
• This work facilitates systematic enhancement of models predicting molecular behavior in different phases.