Chemical Structure Identification from Sparse Mass Spectral Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying the chemical structure of substances using mass spectral data are hindered by irreproducible experimental conditions, limited spectral library sizes, and computational complexity, especially when limited or no mass spectral data is available.
Innovation Solution
A method that involves identifying candidate chemical structure sets using different properties of mass spectral data, determining the similarity between these structures, and using independent searches to increase the likelihood of accurately identifying the substance's chemical structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If library searches are used for identification, then identification speed is improved, but reliability deteriorates due to irreproducible experimental conditions
Solution Approach 1:
The patent transforms the identification approach from direct spectral matching to molecular formula-based searching. By changing the search parameters from requiring identical spectral conditions to accepting molecular formulas derived from mass-to-charge ratios, the method achieves both fast identification and reliability despite experimental condition variations
Solution Approach 2:
The patent introduces molecular formula as an intermediary between the unknown substance and the spectral library. Instead of directly comparing spectra which are sensitive to experimental conditions, the method uses molecular formula (derived from m/z values) as a condition-independent mediator to bridge the unknown substance with candidate structures in the library
2Measurement precision
If spectral library size is increased to improve identification accuracy, then measurement precision is improved, but device complexity worsens
Solution Approach 1:
The patent extracts the essential identification information (molecular formula from m/z ratio) from the complex spectral data, separating the critical identification parameter from the full spectral fingerprint. This allows using a smaller, more manageable library while maintaining or improving identification accuracy
Solution Approach 2:
The patent segments the identification process into distinct steps: first determining molecular formula from m/z values, then searching the library by molecular formula, and finally validating with spectral matching. This segmentation allows using a compact molecular formula index rather than a large spectral library
Data Source
AI summary
A method for analysing mass spectral data of a substance comprises identifying a plurality of candidate chemical structure sets for the substance, each set being identified using one or more respective properties of the mass spectral data. The method comprises determining a level of similarity between candidate chemical structures from different candidate chemical structure sets, so as to determine a likelihood that one of the candidate chemical structures represents the substance.