Exploring molecular mechanisms underlying changes in lipid fingerprinting of salmon (Salmo salar) during air frying integrating machine learning-guided REIMS and lipidomics analysis
Gongshuai Song,Mingwei Zeng,Shengjun Chen,Zhangfan Lyu,Nengliang Jiang,Danli Wang,Tinglan Yuan,Ling Li,Guangming Mei,Qing Shen,Jinyan Gong
DOI: https://doi.org/10.1016/j.foodchem.2024.140770
2024-08-05
Abstract:Lipid oxidation in air-fried seafood poses a risk to human health. However, the effect of a prooxidant environment on lipid oxidation in seafood at different air frying (AF) temperatures remains unknown. An integrated machine learning (ML) - guided REIMS and lipidomics method was applied to explore lipid profiles, lipid oxidation, and lipid metabolic pathways of salmons under different AF temperatures (140, 160, 180, and 200 °C). A significant difference in the lipidomic fingerprinting of air-dried salmon at different temperatures was shown by the main ML methods (neural networks, support vector machines, ensemble learning, and naïve bayes). In total, 773 differential expression metabolites (DEMs) were identified, including glycerophospholipids (GPs), glycerides (GLs), and sphingolipids. A total of 34 DEMs with p values <0.05 and variable importance of projection values >1.0 were analyzed, belonging to linoleic acid metabolism, GL metabolism, and GP metabolism pathways. Correlation network analysis revealed that some characteristic DEMs (phosphatidylcholine, lyso-phosphatidylcholine, triglycerides, fatty acids, and phosphatidylethanolamine) were highly correlated with lipid oxidation. In addition, variations of volatile compounds, color values, texture characteristics, and thiobarbituric acid-reactive substance values were analyzed to corroborate the oxidation characteristics.