Objectives: To discover novel potential biomarkers and establish a diagnostic pattern for SLE by using proteomic technology.
Methods: Serum proteomic spectra were generated by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) combined with weak cationic exchange magnetic beads. A training set of spectra, derived from analysing sera from 32 patients with SLE, 43 patients with other autoimmune diseases and 43 age- and sex-matched healthy volunteers, was used to train and develop a decision tree model with a machine learning algorithm called decision boosting. A blinded testing set, including 32 patients with SLE, 42 patients with other autoimmune diseases and 40 healthy people, was used to determine the accuracy of the model.
Results: The diagnostic pattern with a panel of four potential protein biomarkers of mass-to-charge (m/z) ratio 4070.09, 7770.45, 28 045.1 and 3376.02 could accurately recognize 25 of 32 patients with SLE, 36 of 42 patients with other autoimmune diseases and 36 of 40 healthy people.
Conclusions: The preliminary data suggested a potential application of MALDI-TOF MS combined with magnetic beads as an effective technology to profile serum proteome, and with pattern analysis, a diagnostic model comprising four potential biomarkers was indicated to differentiate individuals with SLE from RA, SS, SSc and healthy controls rapidly and precisely.