Proposed Confidence Scale and ID Score in the Identification of Known-Unknown Compounds Using High Resolution MS Data

Abstract

High-resolution (HR) MS instruments recording HR-full scan allow analysts to go further beyond pre-acquisition choices. Untargeted acquisition can reveal unexpected compounds or concentrations and can be performed for preliminary diagnosis attempt. Then, revealed compounds will have to be identified for interpretations. Whereas the need of reference standards is mandatory to confirm identification, the diverse information collected from HRMS allows identifying unknown compounds with relatively high degree of confidence without reference standards injected in the same analytical sequence. However, there is a necessity to evaluate the degree of confidence in putative identifications, possibly before further targeted analyses. This is why a confidence scale and a score in the identification of (non-peptidic) known-unknown, defined as compounds with entries in database, is proposed for (LC-) HRMS data. The scale is based on two representative documents edited by the European Commission (2007/657/EC) and the Metabolomics Standard Initiative (MSI), in an attempt to build a bridge between the communities of metabolomics and screening labs. With this confidence scale, an identification (ID) score is determined as [a number, a letter, and a number] (e.g., 2D3), from the following three criteria: I, a General Identification Category (1, confirmed, 2, putatively identified, 3, annotated compounds/classes, and 4, unknown); II, a Chromatography Class based on the relative retention time (from the narrowest tolerance, A, to no chromatographic references, D); and III, an Identification Point Level (1, very high, 2, high, and 3, normal level) based on the number of identification points collected. Three putative identification examples of known-unknown will be presented. Graphical Abstract ᅟ.

LC-HRMS Metabolomics for Untargeted Diagnostic Screening in Clinical Laboratories: A Feasibility Study

“…The first reference demands the availability and injection of pure standards, whereas the second gives very general ID categories. Recently, an ID scale and ID score have been proposed to help in the confidence of metabolite identification [ 47 ]. Metabolite identification is usually the most time-consuming step.…”

“…Big databases (>50 × 10 6 entries), such as ChemSpider, PubChem, CAS registry, etc. [ 44 , 45 , 46 ], were not considered here because most compounds registered in these big databases have an extremely low probability to be found in human plasma [ 47 ]. Identification confidence of the putatively identified known-unknown metabolites was evaluated based on a recent publication that proposes the use of an identification (ID) confidence scale and ID score [ 47 ].…”

“…[ 44 , 45 , 46 ], were not considered here because most compounds registered in these big databases have an extremely low probability to be found in human plasma [ 47 ]. Identification confidence of the putatively identified known-unknown metabolites was evaluated based on a recent publication that proposes the use of an identification (ID) confidence scale and ID score [ 47 ]. These ID scores are based on a relative retention time or not, and the collection of ID points from various criteria such as mass accuracy, relative isotopic abundance, fragment ions, fine isotopic distribution, etc.…”

“…These ID scores are based on a relative retention time or not, and the collection of ID points from various criteria such as mass accuracy, relative isotopic abundance, fragment ions, fine isotopic distribution, etc. [ 47 ].…”

Metabolomic study of soft corals from the Colombian Caribbean: PSYCHE and 1H-NMR comparative analysis

“…This article describes a workflow that helped to estimate the confidence levels of compound annotations in reported metabolites that were used as established by the Metabolomics Society at 2017 32 . In summary, this work achieved a confidence level of 2 for features (VIPs) most responsible for cytotoxic activity considering that the feasible structure was compared with data or databases of literature by diagnostic evidence 33 . (Fig.…”