Multiobjective optimization of liquid chromatography-triple-quadrupole mass spectrometry analysis of underivatized human urinary amino acids through chemometrics.

Abstract

Optimization of instrumental settings of a triple-quadrupole mass analyzer was performed by Box-Behnken design, support vector machines, and a Pareto-optimality approach. This time-saving, stepped chemometric strategy was used to model the signal response of underivatized human urinary amino acids. Drying gas flow, nebulizer pressure, sheath gas flow, and capillary voltage settings were exhaustively studied beyond the parameters conventionally optimized in triple-quadrupole devices (multiple reaction monitoring transitions, fragmentor and collision energy voltages). The results indicate that the best signal response for high-abundance and low-abundance underivatized amino acids was achieved with drying gas flow of 9 L/min, nebulizer pressure of 60 psi, sheath gas flow of 13 L/min, and capillary voltage of 3000 V. Compared with the widely standardized settings tested, chemometric analysis led to signal intensities 74% and 68% higher for high-abundance and low-abundance amino acids, respectively. The flexibility, speed, and efficiency of this method allows its affordable implementation in all mass spectrometry-based research to obtain superior results compared with those obtained with conventionally optimized mass spectrometry instrumental parameters.