Modularity and complex interaction in cancer

General info

Date from - to

01 Nov 2009 - 31 Oct 2013

Project leader(s)

Wessels, Lodewyk Dr.

Abstract

Aim of the project:
Identify the players (e.g. genes, pathways) and the interactions between them, which together define clinically relevant groups of cancer patients.

Key objectives:

• Identify modules at different levels of complexity;
• Identify logical interaction networks between these modules;
• Link activity patterns of the networks to phenotypes (response and outcome) on available mouse and human breast cancer cohorts.

Approach:
In this project we will develop a computational framework to perform three tasks:

1. define modules at different levels of complexity (e.g. from genes to pathways to functional modules),
2. model complex interactions between these modules as interaction networks, consisting of basic constructs such as cooperation, redundancy and mutual exclusivity,
3. identify (sub-) interaction graphs whose activity patterns are most strongly associated with clinically relevant phenotypes.

In steps 1 and 2 we will exploit a hybrid of knowledge and data-driven approaches based on both functional annotation data (e.g. GO, KEGG) and measured high-throughput data (expression data, insertional mutagenesis data, copy number data, SNP data and genome-wide protein binding profiles). This computational approach will therefore not only allow us to better map ‘genotypes’ to ‘phenotypes’, but will also shed new light on the cross-scale and within-scale interactions that define these phenotypes, opening up new opportunities for patient stratification, therapy selection and drug discovery.