Research on similarity retrieval method based on mass spectral entropy

Area of Science:

• Analytical Chemistry
• Spectroscopy

Background:

• Compound identification in small molecule research heavily relies on comparing experimental mass spectra against spectral databases.
• Existing methods face challenges due to unequal data lengths, leading to inefficient and inaccurate compound retrieval.
• Current commercial software often employs similarity calculation methods with inherent limitations.

Purpose of the Study:

• To develop novel data processing and similarity calculation methods for mass spectra.
• To address the limitations of existing techniques in handling unequal mass spectral data lengths.
• To improve the accuracy and efficiency of compound identification in mass spectrometry.

Main Methods:

• Proposed two mass spectrometry data processing methods: the 'splicing-filling method' and the 'matching-filling method'.
• Introduced an information entropy-based similarity calculation method for mass spectra.
• Developed an alignment method to convert mass spectra of varying lengths into equal-length representations for accurate similarity computation.

Main Results:

• The alignment methods effectively resolved issues with unequal mass spectral data lengths during similarity calculations.
• Information entropy measurements accurately quantified intensity distribution differences for robust similarity comparisons.
• Example validation demonstrated the reliability and suitability of the mass spectral entropy method for compound identification.

Conclusions:

• The proposed splicing-filling and matching-filling methods provide effective solutions for handling unequal mass spectral data lengths.
• The information entropy-based similarity calculation method offers reliable and accurate results for mass spectral identification.
• These advancements enhance the precision and efficiency of compound identification in small molecule research using mass spectrometry.