Analysis of membrane proteome by data-dependent LC-MS/MS combined with data-independent LC-MSE technique

Proteomics work resembles the search for a needle in a haystack. The identification of protein biomarker requires the removal of the false protein data from the whole protein mixture. For high quality proteomic data, even a strict filtration step using the false discovery rate (FDR) is insufficient for obtaining perfect protein information from the biological samples. In this study, the cyanobacterial whole membrane fraction was applied to the data-dependent analysis (DDA) mode of LC-MS/MS, which was used along with the data-independent LC-MSE technique in order to evaluate the membrane proteomic data. Furthermore, the identified MSE-information (MSE-i) data based on the peptide mass and the retention time were validated by the other database search, i.e., the probability-based MASCOT and de novo search engine PEAKS. In this present study, 208 cyanobacterial proteins with FDR of 5% were identified using the data-independent nano-UPLC/MSE acquisition with the Protein Lynx Global Server (PLGS), and 56 of these proteins were the predicted membrane proteins. When a total of 208 MSE-i proteomic data were applied to the DDA mode of LC-MS/MS, the number of identified membrane proteins was 26 and 33 from MASCOT and PEAKS with a FDR of 5%, respectively. The number of totally overlapped membrane proteins was 25. Therefore, the data-independent LC-MSE identified more proteins with a high confidence.