Statistical profiling for identifying transformation products in an engineered treatment process

This study demonstrated statistical profiling consisting of the analysis of variance (ANOVA) and fold change to efficiently identify transformation products of an organic model compound (i.e., carbamazepine, CBZ) in ozonation. To this end, liquid chromatography (LC)-quadrupole time-of-flight mass spectrometry (QTOF-MS) was employed to measure the accurate masses of CBZ transformation products. Subsequently, statistical profiling was applied to differentiating features that are uniquely present in the ozonated samples from those in blanks and control (i.e., CBZ sample without ozonation). The identified transformation products had significant statistical power (i.e., power, 1-β > 0.8) in post hoc power analysis, which suggests that the profiling procedure can be an efficient means of reducing false negative in data analysis. 2-quinazolinone was newly reported here as a tentative transformation of CBZ during ozonation. In addition, a transformation product with one less carbon than CBZ, often called “anomalous” transformation product, was also found. While statistical profiling was applied to a model experiment, such an approach can be further utilized to screen many features with a higher data complexity such as non-targeted screening (NTS) and non-target analysis (NTA) for environmental samples.