HPLC separations coupled with coulometric electrode-based detectors offer potential to evaluate complex biological samples, but analytical approaches to such samples are still being defined. Coulometric electrode array systems use response ratio (as ratio conformity), for quantitation and qualitative characterization. We investigated the influence of changing the ratio conformity requirement (ratio window, RW) on quantitative consistency of the system. Sixty-one serum metabolites were analyzed in three cohorts of ad libitum/dietary restricted rats with two RW settings. RW 30 uses channels with <30% ratio variation (vs standard) to quantitate, whereas RW 100 generally uses all channels. Quantitation between the two approaches differed in 29/671, 61/976, and 264/976 metabolites in cohorts A, B, and C, respectively. The changes were 23.0 +/- 16.6, 31.9 +/- 28.6, and 23.6 +/- 17.1% (+/- SD). Correlation analysis showed that 11, 7, and 21% of the 61 metabolites had r(2) < 0.9 in cohorts A, B, and C, respectively. Inspection of HPLC chromatograms revealed that most of the metabolites with changed levels possessed common characteristics (e.g., shoulder peaks, low signal/noise ratios, coeluting peaks). Analysis demonstrated that RW 30 provided precise quantitation, whereas RW 100 did not. These data suggest that narrow ratio window increases quantitative accuracy in coulometric electrode array analysis of the metabolome(s) of complex biological samples (e.g., sera, mitochondria).
(C)2002 Elsevier Science (USA).