Genetic algorithm partial least squares (GAPLS) regression for the simultaneous quantitation of cholesterol and mono-/polyunsaturated fatty acids
Gerard G. Dumancas,1 Mary Muriuki,1 Neil Purdie,1 and Lisa Reilly2
1Department of Chemistry, Oklahoma State University, Stillwater, OK, USA 74078
2Department of Physical Sciences, Bethany College, Bethany, WV, USA 26032
Genetic algorithm partial least squares (GAPLS) regression coupled with a patented assay is used for the molar quantitation of eight lipids in human serum. The patented reagent system is selective to the -CH=CH-CH2- group that produces a unique absorbance spectrum for each lipid analyte containing such functionality. In contrast to high performance liquid chromatography (HPLC), gas chromatography-flame ionization detection (GC-FID), or gas chromatography-mass spectrometry (GC-MS), wherein the methods are quite laborious and complicated, no separation steps are necessary. The assay is robust, cheap, and was able to simultaneously quantitate the eight most abundant lipids in human serum consisting of cholesterol, linoleic, linolenic, arachidonic, eicosapentaenoic, docosahexaenoic, conjugated linoleic, and oleic fatty acids in human serum within a 20-minute period using an ultraviolet-visible (UV-VIS) spectrophotometer from 350-550 nm. Partial least squares (PLS1) regression was initially attempted to quantitate the spectra produced from the eight lipids and yielded low root mean square error of prediction (RMSEP) (average of 6.36 µM). Although PLS1 allows one to take into account the whole spectrum without having to perform feature selection, it has been, however, recognized that an efficient feature selection can be highly beneficial both to improve the predictive ability of the model and to greatly reduce its complexity. An attempt was made of further reducing the RMSEP by GAPLS and was successful in its application to almost all lipids analytes. This article shows how GAPLS when coupled to the patented reagent system can simultaneously quantitate the abovementioned eight most abundant lipids within a 20-minute period in synthetic mixtures and human serum. The main advantages of this technology are its appropriateness in a typical clinical setting wherein time, instrumentation, and affordability are the main issues of concern.
|Click [+] for other articles on||The Market Place for Lab Automation & Screening||The Market Place|