Machine learning for identification of silylated derivatives from mass spectra

Abstract

Motivation Compound structure identification is using increasingly more sophisticated computational tools, among which machine learning tools are a recent addition that quickly gains in importance. These tools, of which the method titled Compound Structure Identification:Input Output Kernel Regression (CSI:IOKR) is an excellent example, have been used to elucidate compound structure from mass spectral (MS) data with significant accuracy, confidence and speed. They have, however, largely focused on data coming from liquid chromatography coupled to tandem mass spectrometry (LC–MS). Gas chromatography coupled to mass spectrometry (GC–MS) is an alternative which offers several advantages as compared to LC–MS, including higher data reproducibility. Of special importance is the substantial compound coverage offered by GC–MS, further expanded by derivatization procedures, such as silylation, which can improve the volatility, thermal stability and chromatographic peak shape of semi-volatile analytes. Despite these advantages and the increasing size of compound databases and MS libraries, GC–MS data have not yet been used by machine learning approaches to compound structure identification. Results This study presents a successful application of the CSI:IOKR machine learning method for the identification of environmental contaminants from GC–MS spectra. We use CSI:IOKR as an alternative to exhaustive search of MS libraries, independent of instrumental platform and data processing software. We use a comprehensive dataset of GC–MS spectra of trimethylsilyl derivatives and their molecular structures, derived from a large commercially available MS library, to train a model that maps between spectra and molecular structures. We test the learned model on a different dataset of GC–MS spectra of trimethylsilyl derivatives of environmental contaminants, generated in-house and made publicly available. The results show that 37% (resp. 50%) of the tested compounds are correctly ranked among the top 10 (resp. 20) candidate compounds suggested by the model. Even though spectral comparisons with reference standards or de novo structural elucidations are neccessary to validate the predictions, machine learning provides efficient candidate prioritization and reduction of the time spent for compound annotation. Supplementary Information The online version contains supplementary material available at 10.1186/s13321-022-00636-1.

GC-EI-MS datasets of trimethylsilyl (TMS) and tert-butyl dimethyl silyl (TBDMS) derivatives for development of machine learning-based compound identification approaches

“…Using the given metadata and the described procedure (described in more details by Ljoncheva et al. [2] ), the training GC-EI-MS spectral datasets can be reconstructed.…”

“…Evaluation of the models generated by ML has been conducted separately on spectra of compounds whose spectra appear/do not appear in the training datasets, respectively, and results are reported separately by Ljoncheva et al. [2] . In the metadata files for the testing datasets (“Metadata_test_TMS derivatives.xlsx” and “Metadata_test_TBDMS derivatives.xlsx”), an extra column (the last one) indicates whether a spectrum of the compound at hand also appears in the corresponding training dataset and the respective metadata file (“Metadata_training_TMS_3.3”, “Metadata_ training_TBDMS_3.3”).…”

“…While the training datasets themselves cannot be made publicly available, we make available the corresponding metadata: With licensed address to NIST MSL 17, they can be used to reconstruct the training datasets, by following the workflow summarized below and described by Ljoncheva et al. [2] .…”

“…The ML approach used to build models for the identification of silylated derivatives from these data [2] was the approach titled CSI:OKR [3] .…”

“…The datasets we present here were used to train and test predictive models for identification of silylated derivatives built with ML approaches [4] . The models were built by using data curated from the NIST Mass Spectral Library 17 [2] and the machine learning approach of CSI:Output Kernel Regression (CSI:OKR) [2] . Data from the NIST Mass Spectral Library 17 are commercially available from the National Institute of Standards and Technology (NIST)/U.S.…”