AGEomics Biomarkers and Machine Learning—Realizing the Potential of Protein Glycation in Clinical Diagnostics

Abstract

Protein damage by glycation, oxidation and nitration is a continuous process in the physiological system caused by reactive metabolites associated with dicarbonyl stress, oxidative stress and nitrative stress, respectively. The term AGEomics is defined as multiplexed quantitation of spontaneous modification of proteins damage and other usually low-level modifications associated with a change of structure and function—for example, citrullination and transglutamination. The method of quantitation is stable isotopic dilution analysis liquid chromatography—tandem mass spectrometry (LC-MS/MS). This provides robust quantitation of normal and damaged or modified amino acids concurrently. AGEomics biomarkers have been used in diagnostic algorithms using machine learning methods. In this review, I describe the utility of AGEomics biomarkers and provide evidence why these are close to the phenotype of a condition or disease compared to other metabolites and metabolomic approaches and how to train and test algorithms for clinical diagnostic and screening applications with high accuracy, sensitivity and specificity using machine learning approaches.