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

Area of Science:

• Analytical Chemistry
• Computational Chemistry
• Metabolomics

Background:

• High-throughput identification of unknown metabolites is crucial but challenging.
• Current non-targeted metabolomics primarily uses mass spectrometry, with computational ranking of candidate structures.
• Infrared (IR) spectra have not been extensively evaluated for large-scale metabolite identification.

Purpose of the Study:

• To develop and evaluate a high-throughput computational method for predicting IR spectra of candidate compounds.
• To assess the utility of IR spectra for orthogonal structure discrimination in metabolomics.
• To integrate IR spectra matching with existing mass spectrometry techniques.

Main Methods:

• A computational workflow (IRdentify) was developed, combining fast semiempirical quantum mechanics and density functional theory for IR spectra prediction.
• Candidate structures were sourced from the PubChem database.
• Predicted IR spectra were ranked by similarity to experimental gas-phase IR spectra from NIST.

Main Results:

• The method correctly identified 47% of 258 test compounds.
• An average of 2152 candidate structures were evaluated per test compound.
• The approach demonstrated potential for combining IR and mass spectra, identifying precursor/fragment ions, and analyzing less-volatile compounds.

Conclusions:

• Matching computational and experimental IR spectra offers a powerful orthogonal method for high-throughput chemical structure identification.
• This IR spectra-based approach can significantly enhance discrimination capabilities in non-targeted metabolomics.
• Further applications include composite ranking scores and analysis of derivatized compounds.