Signal Clustering Using Expected Width Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for clustering signals in mass spectra struggle to accurately capture all relevant signals, leading to incorrect data analysis and model formation due to cluster windows that are either too narrow or too wide, which can result in incorrect inclusion or exclusion of signals associated with different markers.
Innovation Solution
The method involves using an expected signal width to define cluster windows, which accounts for the non-linear relation of signal width to time-of-flight or mass-to-charge ratio, allowing for more intuitive clustering and reducing human error by automatically determining the appropriate window size based on empirical or theoretical data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a fixed percentage cluster window is used, then the clustering method is simple to implement, but it fails to account for the non-linear relation of signal width to mass-to-charge ratio, resulting in inaccurate clustering at different mass ranges
Solution Approach 1:
The patent changes the parameter used to define cluster window size from a fixed percentage to an expected signal width value that accounts for the non-linear relationship between signal width and mass-to-charge ratio. This allows the cluster window size to adapt to different mass ranges, improving clustering accuracy while maintaining automated implementation through empirically derived or theoretically calculated expected widths.
2Quantity of substance
If the cluster window is made wider to capture all relevant signals, then more signals are included in clusters, but signals associated with different markers may be incorrectly included
Solution Approach 1:
The patent applies local quality by allowing the cluster window size to vary locally across different mass-to-charge ratio ranges. Instead of using a uniform window size, the expected signal width is determined based on the specific mass range and instrument characteristics, enabling appropriate window sizing for each local region of the spectrum. This ensures that signals are captured accurately without including signals from different markers.
3Measurement precision
If manual tuning of cluster window parameters is performed, then clustering accuracy may be improved, but human error and subjectivity increase
Solution Approach 1:
The patent enables the system to determine expected signal widths automatically through self-service mechanisms. The expected width can be empirically derived from calibration data or theoretically calculated based on instrument parameters, eliminating the need for manual human tuning. This maintains high clustering accuracy while achieving complete automation and reducing human error and subjectivity.
Data Source
AI summary
Methods for processing spectra are disclosed. The method includes obtaining a plurality of spectra, each spectrum in the plurality of spectra comprising a signal including a signal strength as a function of time-of-flight, mass-to-charge ratio, or a value derived from time-of-flight or mass-to-charge ratio. Then, a signal cluster is formed by clustering signals from the plurality of spectra with time-of-flights, mass-to-charge ratios, or values derived from time-of-flights or mass-to-charge ratios that are within a window that is defined using an expected signal width value.