Information Theoretic Mass Spectral Library Search for Comprehensive Two-Dimensional Gas Chromatography with Mass Spectrometry (GCxGC-MS)

摘要

A new information-theoretic compound-identification technique for GCxGC-MS and other MS applications uses a similarity metric between two spectra based on the probability distribution functions of the intensities in the library and the noise in the data. The new search technique works better than NIST-MS Search and PBM techniques for synthetic noise data generated for several noise models. The new search technique works better than NIST MS-Search for real GCxGC-MS data for more than 90percent of the cases. Future work could consider relationships between intensities at multiple mass-to-charge ratios in the library characterization and similarity metric computation. Another direction could extend the library search to work for non-integer mass-to-charge ratios (high-resolution mass spectrometry).