A new method for filling information gaps after peak picking in LC-HRMS data

Non-targeted LC-HRMS is a key analytical technology for the identification of targets, suspects and unknowns and for the profiling of environmental or human samples. While there are many options of peak picking in LC-HRMS extracted ion or whole spectra chromatograms, the heuristic algorithms can not account for all types and shapes of noise that can occur in the data. Even though some algorithms exist to mitigate this problem, most result in many false positives. We introduce a new algorithm that fills in missing peaks while keeping the dataset reliable for evaluation.