Applications of p53 interactome analysis to personalised drug discovery

The stress responsive transcription factor p53 is a powerful tumor suppressor implicated in over 50 % of all human cancers. The complexity and nonlinear dynamics of p53 network coupled with extensive literature is challenging. Systems biology methodologies offer promising tools to investigate large networks, providing structured in silico representations for integrative analysis. The Boolean p53 interactome (PKT206) incorporates the diverse p53 information into a comprehensible framework, and demonstrates good predictive ratios (52 – 71%) using logical steady state analysis. Whilst extensive, Boolean models provide only a qualitative approximation of the system. A prerequisite of diseased in silico models is to accurately represent biological phenomena to characterize network perturbations for effective drug candidates. Thus a quantitative approach is necessary. We have applied a novel signal transduction score flow algorithm (STSFA) to PKT206 for model performance and comparison. The STSFA quantitatively analyses large scale ‘omics’ data, typically not accessible with large networks. STFSA in silico simulations resulted in accurate predictions when compared to various gene expression profiles (P = <1x10), generating a significantly higher proportion of correct predictions than LSSA, (P = 0.003). Furthermore, genes CKS2, WWP1, EPHB4 were identified as prospective drug candidates for osteosarcoma cell treatment by in silico knockout analysis. In summary, refinement of Boolean PKT206 using STFSA has provided a semi quantitative view of the p53 interactome, and along with the use of ‘omics’ data, may be of greater clinical relevance for identification of perturbed pathways and personalized therapies by superimposition of individual genomic profiles.