Repository: Freie Universität Berlin, Math Department

Inferring Proteolytic Processes from Mass Spectrometry Time Series Data Using Degradation Graphs

Aiche, S. and Reinert, K. and Schütte, Ch. and Hildebrand, D. and Schlüter, H. and Conrad, T. O. F. (2012) Inferring Proteolytic Processes from Mass Spectrometry Time Series Data Using Degradation Graphs. PLoS ONE, 7 (7). e40656. ISSN 1932-6203

Full text not available from this repository.

Official URL: http://dx.plos.org/10.1371/journal.pone.0040656

Abstract

Background: Proteases play an essential part in a variety of biological processes. Beside their importance under healthy conditions they are also known to have a crucial role in complex diseases like cancer. It was shown in the last years that not only the fragments produced by proteases but also their dynamics, especially ex vivo, can serve as biomarkers. But so far, only a few approaches were taken to explicitly model the dynamics of proteolysis in the context of mass spectrometry. Results: We introduce a new concept model proteolytic processes, the degradation graph. The degra- dation graph is an extension of the cleavage graph, a data structure to reconstruct and visualize the proteolytic process. In contrast to previous approaches we extended the model to incorporate endoproteolytic processes and present a method to construct a degradation graph from mass spectrometry time-series data. Based on a degradation graph and the intensities extracted from the mass spectra it is possible to estimate reaction rates of the underlying processes. We further suggest a score to rate different degradation graphs in their ability to explain the observed data. This score is used in an iterative heuristic to improve the structure of the initially constructed degradation graph. Conclusion: We show that the proposed method is able to recover all degraded and generated peptides, the underlying reactions, and the reaction rates of proteolytic processes based on mass spectrometry time-series data. We use simulated and real data to demonstrate that a given process can be reconstructed even in the presence of extensive noise, isobaric signals and false identications. While the model is currently only validated on peptide data it is also applicable to proteins, as long as the necessary time series data can be produced.

Item Type:Article
Subjects:Mathematical and Computer Sciences > Mathematics > Mathematical Modelling
Biological Sciences > Molecular Biology > Molecular Biology
Mathematical and Computer Sciences > Computer Science > Computing Science not elsewhere classified
Divisions:Department of Mathematics and Computer Science > Institute of Mathematics > Comp. Proteomics Group
Department of Mathematics and Computer Science > Institute of Mathematics
Department of Mathematics and Computer Science > Institute of Computer Science > Algorithmic Bioinformatics Group
Department of Mathematics and Computer Science > Institute of Computer Science
Department of Mathematics and Computer Science > Institute of Mathematics > BioComputing Group
ID Code:1143
Deposited By: Admin Administrator
Deposited On:11 Jun 2012 17:28
Last Modified:10 Apr 2013 11:52

Repository Staff Only: item control page