Liquid Chromatography Mass Spectrometry (LC-MS) is a powerful and widely applied method for the study of biological systems, biomarker discovery and pharmacological studies. LC-MS measurements are, however, significantly biased by several factors, including: (1) ionization suppression/enhancement, interfering with the correct quantification of analytes; (2) detection of large amounts of derivative ions, increasing the complexity but not the information content; and (3) machine drift during extensive sample sequences, altering mass and quantification accuracy. Traditionally, quality control (QC) samples (e.g., pooled samples of aliquots taken from all samples) are analyzed randomly throughout the analytical sequence. Afterwards the analyst can use the results of the QC samples to determine if there was a gradual change during the analysis or whether a sudden deterioration had occurred at some point midway through the analysis. Based on the analysis of the QC samples it can be decided in the worst case to discard the dataset and repeat the whole experiment, resulting in additional time and money spent.

Here we present a method capable of detecting instrument performance at measurement runtime, offering the possibility of early stop of the whole experiment, saving both time and samples. The method works simply by extracting information from the raw data of a set of analytes known to be present in all samples. These ubiquitous analytes might serve as quality markers. Several mathematical methods are then applied to the RT-values, m/z and intensity data of these analytes, resulting in the detection of significant changes. The method also results in meta-information (achieved mass accuracy, quantification dispersion, adducts, fragments, contaminants, etc.), which will assist in the further identification of unknown metabolites.

The methodology is currently being tested on batch of urine samples measured on a UPLC Q-TOF MS instrument.