Application of A Serum Protein Fingerprinting Model in The Diagnosis of Type 2 Diabetic Nephropathy

Background. The number of cases of diabetic nephropathy is increasing worldwide. Early intervention and consequent therapy may prevent or at least delay the development in these patients. Therefore, a new method is required for diagnosis of diabetic nephropathy at the early stage. Methods: Surface enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS) technology was used to analyze serum proteins of 51 type 2 diabetic nephropathy and 66 controls with CM10 protein chip. A diagnostic model was developed and validated using artificial neural network software. Results: 175 protein peaks were detected at the molecular weight range of 2,000 to 30,000, among which 17 showed significant differences between type 2 diabetic nephropathy and the controls (P<0.01). A diagnostic model was developed and blindly tested by an artificial neural network with 6 proteins. The sensitivity and specificity were 81.0% and 96.2%, respectively. Conclusion: A diagnostic model based on SELDI-TOF-MS and bioinformatics technology is a new approach for the diagnosis of diabetic nephropathy.