A new matching algorithm for high resolution mass spectra.

Area of Science:

• Analytical Chemistry
• Computational Chemistry
• Bioinformatics

Background:

• High-resolution spectra comparison is crucial for chemical analysis.
• Existing spectral matching methods often rely on fixed intervals or alignment, limiting their accuracy.
• Accurate Mass Spectrometry (AMS) generates complex spectral data requiring robust comparison techniques.

Purpose of the Study:

• To introduce a new, alignment-independent algorithm for comparing high-resolution spectra.
• To develop a method for accurate spectral matching using the Jeffreys-Matusitas (JM) distance.
• To enhance spectral database indexing and retrieval performance.

Main Methods:

• Developed an accurate mass spectrum (AMS) distance algorithm based on the Jeffreys-Matusitas (JM) distance.
• The algorithm calculates spectral differences independent of peak alignment and considers varying spectral resolutions.
• Applied the algorithm for indexing a database of 80 accurate mass spectra from Penicillium series Viridicata isolates.

Main Results:

• The AMS distance method demonstrated high retrieval performance, achieving approximately 97-98% accuracy.
• The algorithm proved effective in indexing and retrieving spectra from closely related species.
• The method is independent of variable alignment procedures and binning, simplifying spectral data analysis.

Conclusions:

• The proposed AMS distance algorithm offers a robust and accurate method for high-resolution spectral comparison.
• This alignment-independent approach enhances spectral database usability and retrieval efficiency.
• The method provides a valuable tool for analyzing complex spectral data in fields like microbial identification.