Introduction: High throughput plasma proteomics using data independent acquisition (DIA)is an important approach for biomarker discovery. However, plasma is a formidable matrix to analyse as it is dominated by several high abundance proteins (e.g. albumin, immunoglobulins) that make detection and quantitation of less abundant proteins a challenge. DIA is best used with reference libraries of peptide MS/MS spectra and there are several approach to generate them. In this report we explored the use of reference plasma libraries from public data which had been produced using Orbitrap instruments.
Method: We have adopted a new approach that uses previously acquired MS data from public repositories for peptide MS/MS library generation, effectively negating the time and cost outlay needed for in-house generation. A dataset was downloaded from proteome exchange (Mann Et al PDX00284) and was combined with a single set of high pH reverse phase LCMS data generated on QE-HF instrument, created from an undepleted pool of individual patient samples containing iRT peptides. The final library contained 1344 protein groups, 9966 peptide entries and 2277 proteins. This “enhanced” library was then used as the reference for a pilot study of undepleted plasma from 15 diabetes patients taking fenofibrate. Data was generated on a QE-HF using capillary flow over 60min, with 35 windows over the 350-1200 m/z range.
Results: Fenofibrate is known to cause reduction in inflammatory process associated with diabetes mellitus and increases in lipid metabolism associated with weight reduction. Using a longitudinal analysis of a total of ~250 quantitated proteins, we found that there was a significant up regulation of proteins associated with lipid metabolism and downregulation in proteins associated with inflammation post treatment with fenofibrate. These results confirmed that open source data can be used for library generation for conducting plasma DIA LFQ experiments.