Objective: To screen the serum protein expression profiles in patients having polycystic ovary syndrome (PCOS) with or without insulin resistance (IR) and search for discriminatory proteins.
Method: Fasting serum samples of 30 PCOS patients with IR, 30 PCOS patients without IR, and 30 control individuals from Reproductive Center of Peking University Third Hospital were studied.
Results: There were 27 differential protein peaks between PCOS IR patients and controls, 17 between PCOS non-IR patients and controls, and 19 between PCOS IR patients and non-IR patients. Marker proteins from differentially expressed proteins were screened out using support vector machine (SVM), and were used to establish three diagnostic models for PCOS IR, PCOS non-IR, and IR, respectively.
Conclusions: There were significantly different serum proteomic patterns in different types of PCOS. Using Protein Chip combined with SVM, computer diagnostic models for PCOS with and without IR were set up quickly and efficiently. These discriminatory proteins may help us understand the proteomic changes in serum and find out potential biomarkers of PCOS and IR.