Year: 2026
Type: preprint
Abstract: LC–MS/HRMS-based non-target analysis relies heavily on MS2 spectral matching for compound annotation. However, not all fragments contribute equally to identification confidence, and quantitative measures of fragment ambiguity remain lacking. In this study, we developed a data-driven probabilistic framework to assess fragment level diagnostic power through the Fragment Ambiguity Score (FAS), which combines fragment stability and specificity within precursor-mass-defined candidate spaces. FAS was evaluated using five experimental spectral databases (MassBank, MoNA, GNPS, NIST20, and MSnLib), comprising together of 1,026,855 spectra and an in silico spectral library containing 256,865 spectra. The results showed that MS2 spectra contain fragments with substantially different levels of ambiguity. Using FAS ≤ 0.5 as 12 the threshold for diagnostic fragments, only 14.1% of fragments in the experimental databases and 10.7% in the in silico library were classified as diagnostic. Fragments appearing highly diagnostic in smaller databases often become substantially more ambiguous when evaluated against larger and more diverse spectral collections. Comparison of experimental and in silico spectral databases further illustrated that fragment ambiguity estimates depend on database origin and composition. Application of FAS to suspect screening and spectral library matching demonstrated that fragment level ambiguity provides complementary evidence to spectral similarity. These findings demonstrate that the number of matched fragments alone is insufficient for confidence assessment and highlight the importance of complementary sources of evidence. The proposed framework provides a quantitative basis for assessing fragment ambiguity in spectral databases.
Source: ChemRxiv
Institutions: University of Amsterdam, The University of Queensland
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7204853587