The unavailability of adequate lipid reference standards is a major challenge for accurate quantitative analysis of lipidomics. Based on the discovery of regularity and predictability for lipids in chromatography-mass spectrometry behaviors, "target compound-structure correlation-analytical parameter database" protocol and "modeling-prediction" strategy were carried out to calculate the relative coefficients of analytical parameters within each subclass. Then the relevant LC-tandem-MS parameters of unknown lipids were predicted and a quantification parameter database for 4081 lipids was established and validated. Reference standards-independent accurate determination for lipidomics was achieved with the parameter's database and applied to monitor the change of lipid metabolism in the plasma of whole course of health-hepatitis-cirrhosis-hepatocellular carcinoma (HCC). Combined Student's t test, orthogonal partial least squares discrimination analysis (OPLS-DA) and binary logistic regression-ROC analysis, lipid biomarkers for differentiating health from each disease and differentiating different stages of disease were identified and the pathogenesis of HCC was preliminarily clarified. The established methodology would shed light on comprehensive and accurate quantitative lipidomics and exploring the pathomechanism and potential therapeutic targets of HCC.
Keywords: Hepatocellular carcinoma; Lipidomics; “Modeling-prediction-database” strategy.
Copyright © 2022. Published by Elsevier B.V.