High-Throughput Non-targeted Chemical Structure Identification Using Gas-Phase Infrared Spectra

Abstract

The high-throughput identification of unknown metabolites in biological samples remains challenging. Most current non-targeted metabolomics studies rely on mass spectrometry, followed by computational methods that rank thousands of candidate structures based on how closely their predicted mass spectra match the experimental mass spectrum of an unknown. We reasoned that the infrared (IR) spectra could be used in an analogous manner and could add orthologous structure discrimination; however, this has never been evaluated on large data sets. Here, we present results of a high-throughput computational method for predicting IR spectra of candidate compounds obtained from the PubChem database. Predicted spectra were ranked based on their similarity to gas-phase experimental IR spectra of test compounds obtained from the NIST. Our computational workflow (IRdentify) consists of a fast semiempirical quantum mechanical method for initial IR spectra prediction, ranking, and triaging, followed by a final IR spectra prediction and ranking using density functional theory. This approach resulted in the correct identification of 47% of 258 test compounds. On average, there were 2152 candidate structures evaluated for each test compound, giving a total of approximately 555,200 candidate structures evaluated. We discuss several variables that influenced the identification accuracy and then demonstrate the potential application of this approach in three areas: (1) combining IR and mass spectra rankings into a single composite rank score, (2) identifying the precursor and fragment ions using cryogenic ion vibrational spectroscopy, and (3) the incorporation of a trimethylsilyl derivatization step to extend the method compatibility to less-volatile compounds. Overall, our results suggest that matching computational with experimental IR spectra is a potentially powerful orthogonal option for adding significant highthroughput chemical structure discrimination when used with other non-targeted chemical structure identification methods.

Combining Experimental with Computational Infrared and Mass Spectra for High-Throughput Nontargeted Chemical Structure Identification

“…This relatively low false negative rate is much better than the nearly 7% false negative rate we found in our earlier work using a molecular mechanics IR prediction method for triaging. 63 Additionally, CSI:FingerID scoring and ranking accuracy were not significantly affected by the size of the candidate list (Figure S3a). This is important since some candidate lists included more than 15,000 candidate structures (Figure S1c).…”

“…Importantly, the triage cutoff level we selected (≤1000 compounds) resulted in incorrectly eliminating only 3 of 148 (2%) test compounds. This relatively low false negative rate is much better than the nearly 7% false negative rate we found in our earlier work using a molecular mechanics IR prediction method for triaging . Additionally, CSI:FingerID scoring and ranking accuracy were not significantly affected by the size of the candidate list (Figure S3a).…”

“…For both of these methods, computational predicted IR spectra compared favorably with experimental data. We previously showed that by matching de novo computational predictions of IR spectra with experimental IR spectra, we could obtain structure identification rates comparable to using MS/MS data alone . However, the combined use of MS/MS and IR experimental and computational methods has not been reported for large data sets.…”

“…This resulted in an average of 4444 candidate structures scored and ranked for each test compound for a total of 657,712 candidate structures evaluated. We reasoned that combining CSI:FingerID with our previously reported IR-based method would be potentially advantageous because not only these analytical methods are orthologous but also CSI:FingerID employs machine learning for scoring and ranking candidate structures and is significantly faster than the IR-dentify method used previously for scoring and ranking candidates . Thus, this combined approach would potentially improve identification accuracy and reliability while at the same time increasing computational speed.…”

An In Silico Infrared Spectral Library of Molecular Ions for Metabolite Identification

“…Using quantum-chemical methods, vibrational spectra can be routinely and accurately calculated (somewhat in contrast to MS/MS spectra). For example, Karunaratne et al pursued identification of neutral gaseous molecules detected with GC-FTIR spectroscopy by employing DFT-calculated reference spectra of compounds extracted from the PubChem database . This approach yields identification rates comparable to those achieved with MS/MS.…”