Integration, analysis and meta-mining of data to aid disease diagnosis, prognosis and treatment response
Session type: Parallel sessions
1Breakthrough Breast Cancer Research, London, UK, 2Institute for Cancer Research, London, UK
Proffered paper presentation
In systems biology, biologically relevant quantitative modelling of physiological processes requires the integration of experimental data from diverse sources. Recent developments in high-throughput methodologies enable analysis on a previously unprecedented scale, causing a deluge of data in public databases. Effective integration and analysis of this data can lead to better understanding and treatment of breast cancer with the minimum of side effects.However, a key requirement of this analysis is the understanding of the networks of interactions between cellular components.
A major step in building this understanding is the combinatorial analysis of data from different experimental sources. This requires not only a wide range of experimental approaches, but also a central, integrated database and associated analysis and data-mining tools. For example, microarray technology has enabled the development of a range of genome-scale analyses, such as analysis of DNA copy number (aCGH), promoter methylation, SNP genotyping and gene expression. We have developed an online database with associated analysis and data-mining tools to facilitate integration of RNAi, aCGH and gene expression data with networks and pathways.We can also cross-correlate SNP genotyping, methylation, drugs/ligand and protein structure data as well as Next Generation Sequencing to facilitate the detailed mapping of key pathways involved in breast cancer.
Combinatorial analysis and integration with network and pathway information is rapidly emerging as a powerful technique to identify the most biologically significant results from these large, noisy datasets.
This integration and mathematical modelling approach should aid disease diagnosis, prognosis and treatment response.