Enhanced authentication of organic milk using MALDI-TOF MS with combined lipid-peptide fingerprinting and machine learning integration

This study introduces a method for authenticating organic milk and determining its geographic region using matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS). Both lipids and peptides were analyzed, and their combined dataset significantly improved classification accuracy when integrated with machine learning models. To improve method reproducibility, a binary matrix composed of alpha-cyano4-hydroxycinnamic acid and 2,5-dihydroxybenzoic acid, combined with a vacuum drying strategy, was implemented. This approach reduced the relative standard deviation by approximately 25 % compared to traditional matrices. Multivariate statistical analysis identified key lipids and peptides as markers for discriminating organic from conventional milk and for geographic region classification. Integrating the lipid-peptide dataset with machine learning achieved up to 100 % accuracy, outperforming models based on single-dataset. Additionally, twoway ANOVA revealed that farming system, geographic region and their interactions significantly influenced milk composition. This method demonstrates the potential of combining MALDI-TOF MS with machine learning to prevent organic milk fraud.