Detection and correction of interference in mass spectrometry quantitation of peptides with potential application to nuclear pore complex studies

In the past decade, mass spectrometry (MS) has been the technology of choice for accurate quantitation of selected peptides. However, the frequent occurrence of interference is a significant problem which causes false-positive identification and inaccurate quantitation of peptides. We present here approaches to detect and correct interference in both MS1 and MS2 quantitation. The natural isotopic abundance of ions in a peptide is a property independent of peptide concentration; therefore the interference in MS1 quantitation of peptides is detected using peptide isotope distributions through our approach. We also use computational simulations to build a library of possible MS1 measurements with interference. We correct the quantitation of peptides with detected interference by searching the simulation library. This approach is tested with E. coli and S. cerevisiae peptide mixtures. For quantitation performed using MS2, we focus on Multiple Reaction Monitoring (MRM) method. We use the expected relative intensity of MRM transitions to detect and correct interference. We design an algorithm to automatically detect the linear range of a calibration curve. We apply our approach to the experimental data of Clinical Proteomic Tumor Analysis Consortium (CPTAC) Verification Work Group Study 7 and show that the corrected measurements provide more accurate quantitation than the uncorrected data. We conduct studies on Nuclear Pore Complex, which require intense mass spectrometry based quantitation techniques involved, and we can apply our interference detection and correction approach to improve the quantitation. We focus on the Nup107 and Nup93 subcomplex, and the pore membrane protein NDC1. With HEK293 cells over-expressing N-terminally His-Strep-HA-tagged Nups, we successfully induced the expression of tags and isolated the members of the Nup107 and Nup93 subcomplexes, which will facilitate crosslinking and structural studies. We also present the initial result about the characterization of a spliced variant of NDC1, which is lacking the transmembrane domains.