Evolving neural network optimization of cholesteryl ester separation by reversed-phase HPLC.

Area of Science:

• Analytical Chemistry
• Biochemistry
• Immunology

Background:

• Cholesteryl esters possess antimicrobial properties, contributing to innate immunity.
• Characterizing cholesteryl esters requires advanced separation techniques.
• Existing methods face challenges in resolving complex mixtures.

Purpose of the Study:

• To optimize reversed-phase high-performance liquid chromatography (RP-HPLC) for cholesteryl ester separation.
• To apply artificial neural network-genetic algorithm (ANN-GA) modeling for chromatographic method development.
• To improve the analysis of cholesteryl esters in biological fluids.

Main Methods:

• Utilized a fractional factorial design to assess factors like mobile phase composition, column temperature, and flow rate.
• Employed artificial neural network-genetic algorithm (ANN-GA) modeling for optimization.
• Integrated Derringer's desirability function to merge separation parameters for model training.

Main Results:

• Achieved complete separation of six analytes, including previously co-eluting cholesteryl esters.
• Validated the ANN-GA model with experimental data showing good agreement.
• Demonstrated successful application of the optimized method to human milk samples.

Conclusions:

• ANN-GA modeling effectively optimizes chromatographic separation of cholesteryl esters.
• The developed method enhances the resolution of complex biological samples.
• This approach provides a valuable tool for future research on cholesteryl esters and innate immunity.