P109 Construction of a suite of computable biological network models focused on inflammatory bowel disease
P.A. Ruiz Castro, H. Yepiskoposyan, J. Szostak, M. Talikka, J. Hoeng
PMI, Systems toxicology, Neuchatel, Switzerland
Background
Inflammatory bowel diseases (IBDs) are chronic relapsing inflammatory disorders that result from an inappropriate inflammatory response to the enteric microbiota in a genetically susceptible host. Because the molecular mechanisms of IBD have been the subject of intensive exploration, we assembled the available information into a suite of comprehensive causal biological network models, which offer a better visualisation and understanding of the processes underlying IBD.
Methods
Biological expression language (BEL), which represents biological findings in a computable form was used for scientific text curation. The curated BEL statements were compiled using the OpenBEL framework 3.0.0. Cytoscape was used to visualise and analyse network properties. The network perturbation amplitude methodology was used to score the network models with transcriptomic data from public data repositories.
Results
The IBD network model suite consists of four independent models that represent signalling pathways contributing to biological processes involved in IBD. In the ‘wound healing’ model, fibroblast growth factor and thymosin β-4-associated signalling pathways emerged as the main routes leading to epithelial cell proliferation and thus to wound repair. Phosphoinositide 3-kinases and carcinoembryonic antigen-related cell adhesion molecule-associated pathways were identified in the ‘bacterial colonisation’ model as the primary pathways modulating bacterial adhesion and internalisation. In the ‘immunity’ model, nuclear factor-κB and myeloid differentiation primary response 88 emerged as the main contributors to immune responses. In the ‘epithelial barrier defence’ model, tumour necrosis factor- and interleukin 1-associated signalling cascades were the main routes leading to tight junction disruption and thereby, to increased permeability in the intestinal epithelium. The scoring of publicly available transcriptomic data sets onto these network models demonstrates that the IBD models capture the perturbation in each dataset accurately.
Conclusion
This work is a continuation of our approach using computational biological network models and mathematical algorithms in order to interpret high-throughput molecular datasets. The IBD network model suite can provide better mechanistic insights of the transcriptional changes in IBD and constitutes a valuable tool in personalised medicine to further understand molecular heterogeneity and individual drug responses in IBD.