LC-MS Lipid Identification via Deconvolution and Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current LC-MS-based methods for lipid identification are exploratory and lack a systematic approach for prioritizing good leads while maintaining false ones at a low level, relying heavily on user expertise and experience, and are limited by the availability of rigorous retention time data, which is not consistently available for all system-specific metabolites.
Innovation Solution
A method involving deconvolution of LC-MS-based mass features, inference of daughter ions, scoring of parental exact masses, and determination of characteristic mass features to identify lipids, which includes steps like providing a list of LC-MS-based mass features, deconvoluting them, inferring daughter ions, identifying parental exact masses, scoring each, and determining the lipids based on characteristic mass features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If LC-MS-based exploratory methods are used for lipid identification, then coverage of lipid species can be expanded, but reliability of identification decreases due to lack of systematic prioritization and high false candidate rates
Solution Approach 1:
The patent applies parameter changes by systematically varying scoring parameters (e.g., mass accuracy thresholds, intensity ratios, retention time windows) to prioritize lipid candidates. This allows the system to adjust identification stringency dynamically, improving reliability while maintaining broad coverage through optimized parameter sets for different lipid classes.
Solution Approach 2:
The patent implements feedback mechanisms where identification results from initial screening inform subsequent validation steps. Candidates are scored and ranked based on multiple criteria, with top-scoring candidates prioritized for confirmatory analysis. This feedback loop systematically reduces false positives while maintaining high sensitivity for detecting diverse lipid species.
2Reliability
If user expertise and experience are relied upon for candidate selection, then identification accuracy may improve, but ease of operation decreases and workflow becomes biased
Solution Approach 1:
The patent enables self-service by providing automated, expert-level lipid identification capabilities accessible to users without specialized metabolomics expertise. The system incorporates built-in scoring algorithms, database matching, and validation workflows that automatically perform functions previously requiring expert manual intervention, thereby improving ease of operation while maintaining high identification accuracy.
Solution Approach 2:
The patent replaces manual expert analysis (mechanical human cognition) with automated computational systems. Machine learning algorithms and automated scoring systems substitute for human expert judgment in prioritizing lipid candidates, eliminating workflow bias while maintaining or improving identification accuracy through consistent, reproducible automated decision-making.
3Measurement precision
If retention time filtering is used to improve confidence in lipid identifications, then measurement precision increases, but productivity decreases due to unavailable RT data for system-specific metabolites
Solution Approach 1:
The patent applies universality by developing a multi-functional identification framework that works with or without retention time data. The system can operate in RT-dependent mode when RT data is available (higher confidence) or RT-independent mode using alternative criteria like mass accuracy, intensity ratios, and spectral matching (maintains productivity). This universal approach ensures high identification confidence across diverse metabolite types regardless of RT availability.
4Reliability
If comprehensive validation using MS2 or spectral comparison is performed, then reliability of lipid identification improves, but productivity and time consumption increase
Solution Approach 1:
The patent applies partial action by implementing a tiered validation strategy where not all lipid candidates undergo exhaustive MS2 or spectral comparison validation. Instead, candidates are prioritized based on initial scoring, and only top-scoring candidates requiring confirmation receive comprehensive validation. This selective approach maintains high reliability for confirmed identifications while preserving productivity by avoiding unnecessary validation of low-priority candidates.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method enhances the accuracy and coverage of lipid identification, particularly for untargeted new species, reduces false identifications, and allows for high-throughput analysis without requiring deep expertise in metabolomics or lipidomics, enabling automated and robust lipid profiling.
Implementation Method 1
the primary data obtained are the mass-to-charge (m/z) ratio, intensity, and retention time (RT) of detected mass features representing products of MS ionization
Implementation Method 2
Lipids are generally identified using combined liquid chromatography-mass spectroscopy (LC-MS) techniques due to their wide coverage in terms of molecular weight, ease of sample preparation and tunable column chemistry and mobile phases
Implementation Method 3
ions-of-interest are selected from the first stage of MS based on m/z values, for fragmentation by collision-induced dissociations
Data Source
AI summary
The present invention generally relates to a method for analyzing and identification of the plurality of lipids in a sample that is profiled using a combined Liquid Chromatography-Mass Spectrometry (LC-MS) technique, comprising the steps of:a) providing a list of Liquid Chromatography-Mass Spectrometry (LC-MS)-based mass features;b) deconvoluting said list of LC-MS-based mass features;c) inferring daughter ions from the deconvoluted list of LC-MS-based mass features;d) identifying one or more parental exact masses from the inferred daughter ions;e) scoring each of the one more parental exact masses based on the inferred daughter ions;f) determining characteristic mass features in response to the scoring of each of the one or more parental exact masses; andg) determining each of the plurality of lipids based on the characteristic mass features thereof.In particular, the present invention also relates to identification of the plurality of lipids undergoing in-source fragmentation.