The gold standard for unambiguous identification of a small molecule in metabolomics is based on comparing two or more orthogonal properties of data from analysis of reference materials to experimental data acquired in the same laboratory with the same analytical methods. This represents a significant limitation since many molecules are not available as standard reference material. To move towards standards-free small molecule identification, we are (i) advancing chemical property predictions (e.g. NMR and IR spectra prediction, collision cross section) through the creation and utilization of a large-scale computational chemistry platform, the <i>in silico</i> chemical library engine (ISiCLE), and (ii) developing multi-feature scoring and matching algorithms. To test our methods, we participated in a blinded analysis of synthetic mixtures composed of EPA ToxCast library compounds as part of a 24-laboratory challenge initiated by the U.S. Environmental Protection Agency (EPA). Accurate mass, isotopic signature, and collision cross section (CCS) measurements were collected using ion mobility spectrometry–mass spectrometry (IMS-MS) and ultra-high resolution 21-Tesla Fourier transform ion cyclotron resonance–mass spectrometry (FTICR-MS). Properties were also calculated for each compound in the EPA ToxCast library, allowing us to match observed features to library entries. Our blinded accuracy and false discovery rate for identifications was 91% and 77%. The false discovery rate for “highly confident” identifications was 35%. Additionally, we observed a high number of compounds from the library that were not spiked in but possible through unintentional transformations in the highly concentrated mixtures. Finally, we improved our approach using Monte Carlo to optimize scoring parameters, increasing our accuracy and number of identifications.

Standards-free identification of small molecules using multi-feature matching

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.