Knowledge-based network analysis of genomic data

Introduction

Extracting new biological insight from high throughput genomic studies of human diseases is a daunting challenge, limited by difficulties in recognizing and evaluating relevant biological processes from the immense quantities of experimental data. Cluster and principal component analyses describe overall changes in apparent gene expression, but provide few insights into the biological processes and signalling networks invoked in propagation and resolution of the inflammatory response. Identifying the perturbed biological networks underlying this complex clinical phenotype requires systematic analysis in the context of known mammalian biology, derived from basic and clinical research.

In collaboration with Ingenuity Inc., we developed a structured network knowledge-base approach to analyze the genome-wide transcriptional response in the context of known functional interrelationships among proteins, small molecules and phenotypes. Briefly, using a web-based entry tool developed by Ingenuity, findings presented in peer-reviewed scientific publications were systematically encoded into an ontology by content and modelling experts. Using over 200,000 full-text scientific articles, a knowledge base of more than 9,800 human, 7,900 mouse and 5,000 rat genes was manually curated and supplemented with curated relationships parsed from MEDLINE abstracts. A molecular network of direct physical, transcriptional and enzymatic interactions observed between mammalian orthologues—the observed ‘interactome’—was computed from this knowledge base. 

The observed interactome provides a framework for structuring the existing knowledge regarding mammalian biology, and enables a new analytical approach that objectively examines experimental data in the context of known genome-wide interactions in order to identify significant functional modules. This method is applicable to data of high-throughput platforms such as gene expression profiling, polymorphism analysis and proteomics. Briefly, the data will be overlaid on the observed human interactome in, and be systematically examined for (1) known signal transduction pathways and their expected transcriptional effect on downstream target genes; (2) known genome-wide transcriptional networks to identify key regulatory points of the observed transcriptome; (3) known metabolic networks to identify metabolic reactions and pathways significantly changed; and (4) known proteinprotein complexes (transient and stable) to identify significant complexes. This methodology was utilized to analyze the genomic response to inflammation in human patients and explores the known genome-wide interaction network to identify significant functional modules perturbed in response to acute inflammatory stress.

 

 

References 

Calvano, S.E.*, Xiao, W.*, Richards, D.R., Felciano, R.M., Baker, H.V., Cho, R.J., Chen, R.O., Brownstein, B.H., Cobb, J.P., Tschoeke, S.K., et al. (2005). A network-based analysis of systemic inflammation in humans. Nature 437, 1032-1037.

http://www.gluegrant.org/pubsupport/Nature_1/ 

Laudanski, K.*, Miller-Graziano, C.*, Xiao, W.*, Mindrinos, M.N., Richards, D.R., De, A., Moldawer, L.L., Maier, R.V., Bankey, P., Baker, H.V., et al. (2006). Cell-specific expression and pathway analyses reveal alterations in trauma-related human T cell and monocyte pathways. Proc Natl Acad Sci U S A 103, 15564-15569.

http://www.gluegrant.org/pubsupport/supplement-3/ 

Leave a Reply

Your email address will not be published. Required fields are marked *