A Novel Differential Analysis Algorithm for Low/High Resolution LC/MS Data: Applied to the Detection of Drug Metabolites and GSH-trapped Adducts

摘要

Early metabolite identification provides essential information in synthesizing metabolically stable compounds and identifying pharmacologically active or toxic metabolites. Due to interferences from biological matrices, identification has primarily relied on prediction of expected metabolite masses, mass defects and fragmentation patterns. Such an approach limits the detection of all types of metabolites and adducts. To eliminate false-positive peaks or endogenous interferences, a fast Differential Analyals between sample and control allows for a substantial improvement in throughput of data processing for metabolite profiling studies. Proper chromatographic alignment between sample and control is one of the prerequisites for successful background correction. We present a fast algorithm to detect significant differences between sample and control that operates on both low and high resolution LC/MS data. Differential Analysis is applied on the raw or pre-processed data after peak picking of all significant chromatographic peaks. The end result will be a table listing all significant peaks not present in the control sample.