Modelling pathways to Rubisco degradation: a structural equation network modelling approach

• Published in: PloS one 2014-02, Vol.9 (2), p.e87597-e87597
• Main Authors: Tétard-Jones, Catherine, Gatehouse, Angharad M R, Cooper, Julia, Leifert, Carlo, Rushton, Steven
• Format: Article
• Language: English
• Subjects: Amino Acid Sequence; Arthritis; Bayes Theorem; Bayesian analysis; Biology; Chloroplasts; Computer simulation; Degradation; Degradation products; Efficiency; Epidemiology; Farming; Gene expression; Gene Expression Profiling - methods; Gene Expression Regulation, Enzymologic; Markov Chains; Markov processes; Mass spectrometry; Mass spectroscopy; Mathematics; Metabolic Networks and Pathways; Modelling; Models, Biological; Molecular biology; Molecular Sequence Data; Monte Carlo Method; Monte Carlo methods; Monte Carlo simulation; Nitrogen; Physiology; Protein interaction; Protein Subunits - genetics; Protein Subunits - metabolism; Protein-protein interactions; Proteins; Proteolysis; Proteomics; Proteomics - methods; Reproducibility of Results; Ribulose-bisphosphate carboxylase; Ribulose-Bisphosphate Carboxylase - genetics; Ribulose-Bisphosphate Carboxylase - metabolism; Rural development; Senescence; Statistical analysis; Studies; Two dimensional analysis; Wheat
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0087597

'Omics analysis (transcriptomics, proteomics) quantifies changes in gene/protein expression, providing a snapshot of changes in biochemical pathways over time. Although tools such as modelling that are needed to investigate the relationships between genes/proteins already exist, they are rarely utilised. We consider the potential for using Structural Equation Modelling to investigate protein-protein interactions in a proposed Rubisco protein degradation pathway using previously published data from 2D electrophoresis and mass spectrometry proteome analysis. These informed the development of a prior model that hypothesised a pathway of Rubisco Large Subunit and Small Subunit degradation, producing both primary and secondary degradation products. While some of the putative pathways were confirmed by the modelling approach, the model also demonstrated features that had not been originally hypothesised. We used Bayesian analysis based on Markov Chain Monte Carlo simulation to generate output statistics suggesting that the model had replicated the variation in the observed data due to protein-protein interactions. This study represents an early step in the development of approaches that seek to enable the full utilisation of information regarding the dynamics of biochemical pathways contained within proteomics data. As these approaches gain attention, they will guide the design and conduct of experiments that enable 'Omics modelling to become a common place practice within molecular biology.