New Library-Based Methods for Nontargeted Compound Identification by GC-EI-MS.

Area of Science:

• Analytical Chemistry
• Environmental Science

Background:

• Gas chromatography-mass spectrometry (GC-MS) is crucial for identifying compounds in complex mixtures.
• Current GC-MS identification processes are often subjective, time-consuming, and result in many unidentified spectra.

Purpose of the Study:

• To develop and evaluate new mass spectral library-based methods for enhanced compound identification.
• To address the challenge of unidentified spectra in complex mixtures, particularly from combustion sources.

Main Methods:

• Utilized mass spectral uniqueness, compound ubiquity, noise reduction, and retention index comparison.
• Employed the NIST 2023 EI-MS Library and NIST MS PepSearch software with retention-index corrected Identity Search scoring.
• Introduced Hybrid Similarity Search for identifying related compounds not directly in libraries and developed a median relative abundance measure.

Main Results:

• Initial identification using NIST software was performed on a dataset of 4833 spectra from wildland fuel combustion.
• Despite advanced methods, approximately 90% of spectra remained unidentified.
• A new metric, median relative abundance, was developed to assess the probability of spectral identification.

Conclusions:

• The developed methods offer improvements for compound identification in complex mixtures.
• The median relative abundance metric provides a valuable tool for evaluating the confidence of spectral identifications.
• Further research is needed to address the high percentage of unidentified spectra in challenging samples.