Assessment of the Yeast Proteome Repertoire Using the Mascot Algorithm

Abstract

Recent studies have indicated that transcription and translation are more pervasive on a genome-wide level than previously thought (Xu et al., 2009, Nature 457: 1033; Ingolia et al., 2009 Science 324:218). Specifically, studies using ribosome profiling have shown evidence of translation at previously un-annotated open reading frames (ORFs). These findings challenge current approaches to genome annotation. As a result, new methods are being developed in order to more precisely visualize the proteome. One methodology is peptide mass spectrometry wherein we perform tandem MS/MS on whole cell lysate and use a peptide search algorithm to match theoretical peptides provided by the user to observed mass spectra. To test the accuracy of these algorithms, we developed a gel slice method for parent-protein profiling in order to assess the accuracy of these algorithms (Lin et al., 2014, J. Prot. Res. 13: 1823). In Chapter 1, we used this methodology to assess the performance of the Mascot search algorithm. In Chapter 2, we applied the findings of Chapter 1 in order to assess a previously examined set of novel peptides resulting from the translation of open reading frames (ORFs) downstream of the annotated ORF (dnORFs) (Fournier et al., 2012, J. Prot. Res 11: 5712). In Chapter 3, we applied the methodologies described in Chapters 1 and 2 to detect peptides resulting from the translation of long non-coding RNAs (lncRNAs). In summary, the use of peptide mass spectrometry with peptide search algorithms can provide a high-confidence VI assessment of the proteome. This enabled us to detect previously un-annotated proteins and begin to characterize them.