Soft-trilinear constraints for improved quantitation in multivariate curve resolution

Area of Science:

• Analytical Chemistry
• Chemometrics
• Spectroscopy

Background:

• Hyphenated techniques like GC/MS and LC/MS generate complex data matrices.
• Self-Modeling Curve Resolution (SMCR) methods assume trilinear data structure for unique peak resolution.
• Real-world data often exhibit non-trilinear behavior due to matrix effects or instrumental variations.

Purpose of the Study:

• To introduce and evaluate "soft-trilinearity constraints" for SMCR.
• To address limitations of strict trilinearity assumptions in analyzing non-ideal chemical data.
• To compare the performance of soft-trilinearity constraints against other methods like PARAFAC2.

Main Methods:

• Development of soft-trilinearity constraints within SMCR algorithms.
• Application to a simulated 3-component system.
• Validation using an experimental dataset.
• Comparison with PARAFAC2 and non-negativity constraints.

Main Results:

• Soft-trilinearity constraints allow for small deviations in peak shape and position across samples.
• This approach significantly reduces the range of possible solutions compared to non-negativity constraints.
• Strict trilinearity constraints can lead to inaccurate or impossible solutions in non-trilinear systems.

Conclusions:

• Soft-trilinearity constraints offer a more robust approach for SMCR in the presence of real-world data deviations.
• This method enhances the accuracy and reliability of resolving overlapping peaks in complex mixtures.
• The findings highlight the importance of flexible constraints in chemometric data analysis.