Estimating actual false localization rates for large-scale proteomics datasets / Thomas Taus
VerfasserTaus, Thomas
Begutachter / BegutachterinMach, Robert
UmfangVI, 64 Bl. : graph. Darst.
HochschulschriftWien, Techn. Univ., Dipl.-Arb., 2014
Zsfassung in dt. Sprache
Schlagwörter (DE)Protein Phosphorylierung / Massenspektrometrie
Schlagwörter (EN)protein phosphorylation mass spectroscopy
URNurn:nbn:at:at-ubtuw:1-102410 Persistent Identifier (URN)
 Das Werk ist frei verfügbar
Estimating actual false localization rates for large-scale proteomics datasets [3.36 mb]
Zusammenfassung (Englisch)

In recent years, mass spectrometry has emerged as the technology of choice for protein characterization, including the analysis of post-translational modifications (PTMs), such as phosphorylation. Besides identification and quantification of modified proteins and peptides, the precise localization of the exact site within the amino acid sequence is critical for the biological questions addressed. A variety of software tools have been introduced that determine the most probable PTM assignment and provide scores estimating the confidence of reported results, thereby enabling high-throughput site localization. However, the false localization rate (FLR) of obtained sites cannot yet be estimated accurately by a generally accepted approach and, thus, it was the aim to develop such a statistical method.The novel approach involves addition of proper candidates that are a priori known to be incorrect (decoys) to the search space. Generation of decoy sites is achieved by m/z shifting of site determining fragment ions. In order to validate the method, a peptide library was constructed that comprises roughly 60,000 distinct phospho-peptides. Based on more than 700,000 high confident peptide to spectrum matches that were acquired with diverse dissociation methods and measured under various instrumental conditions, it is assumed that the novel approach is capable of reliable FLR estimation in all scenarios under investigation.