Optimization and Validation of High-Resolution Mass Spectrometry Data Analysis Parameters.

High-resolution mass spectrometry (HRMS) has gained recognition as a valuable tool for comprehensive drug screening in a variety of biological matrices. HRMS instruments collect untargeted, accurate mass data, which permit identification of known and unknown compounds in a single analytical run. One of the most challenging aspects of implementing an HRMS drug screen is establishing appropriate data analysis parameters for identifying compounds. Unlike other types of mass spectrometry data, guidelines for HRMS data analysis and acceptability criteria have not been established. Although many laboratories have published on the utility of HRMS for drug screening, few have included details on how they determined allowable errors and set positivity criteria. Previously, we developed and validated a comprehensive 169-compound drug screen on a high-resolution quadrupole time of flight mass spectrometer. Here we report the detailed procedure that we used to determine appropriate positivity criteria for our screening procedure. Our approach was empirical; we collected data and analyzed it with commonly available software. We found that a combined scoring approach using a threshold of 70, with 70% weight given to library match and 10% weight given to each of mass error, retention time error and isotope pattern difference provided optimum drug identification efficiency of 99.2%. Our results demonstrate the importance of library matching in accurately identifying compounds, and underscore the utility of robust product ion spectra that contain information on the lineage, mass and relative abundance of fragments. The method we describe is easily adaptable to include alternative parameters that may be available in software associated with a variety of HRMS platforms. With careful selection of error limits and positivity criteria, HRMS instruments are capable of producing high-quality, high-confidence results that may reduce the need for confirmatory testing.