Signal Response Evaluation Applied to Untargeted Mass Spectrometry Data to Improve Data Interpretability

Feature finding is a common way to process untargetedmass spectrometry(MS) data to obtain a list of chemicals present in a sample. Mostfeature finding algorithms nai''vely search for patterns of uniquedescriptors (e.g., m/z, retention time, and mobility) and provide a list of unannotatedfeatures. There is a need for solutions in processing untargeted MSdata, independent of chemical or origin, to assess features basedon measurement quality with the aim of improving interpretation. Here,we report the signal response evaluation as a method by which to assessthe individual features observed in untargeted MS data. The basisof this method is the ubiquitous relationship between the amount andresponse in all MS measurements. Three different metrics with user-definedparameters can be used to assess the monotonic or linear relationshipof each feature in a dilution series or multiple injection volumes.We demonstrate this approach in metabolomics data obtained from auniform biological matrix (NIST SRM 1950) and a variable biologicalmatrix (murine kidney tissue). The code is provided to facilitate implementation of this data processing method.

Automated summary

Feature finding is a common method in processing untargeted mass spectrometry data to obtain a list of chemicals present in a sample. This study introduces signal response evaluation as a method to assess the individual features observed in untargeted MS data.