To Develop New Methods That Increase the Power of High-Throughput Experiments By

Introduction

In recent years, several studies have been initiated within VLAG and NuGO (the European nutrigenomics organization) that combine high-throughput transcriptomics and proteomics techniques. Both techniques have their advantages and disadvantages. Transcriptomics data is more mature (in particular microarray data) and more dedicated analysis methods have been developed, but proteomics data better approximates the actual biological process under study. Therefore an integrated approach is needed.

This integration can take two forms. One approach is incorporating understanding from previous research by using pathway analysis. In pathway analysis the data can be visualized and summarized in the context of biological pathways. However, the usefulness of this method heavily depends on the quality of the biological pathways which again depends on the speed with which research findings are incorporated in available pathway repositories.

Another approach is to combine data from different sources to improve power in statistical methods such as Gene Set Enrichment Analysis. This can be especially advantageous for nutritional studies where responses to experimental changes are often subtle and complex, because organisms generally try to maintain the steady state in spite of changes in nutrition.

Aim

To develop new methods that increase the power of high-throughput experiments by

combining proteomics and transcriptomics data in the context of biological pathways. This can be divided in three parts: (1) To visually integrate proteomics and transcriptomics data in the context of biological pathways. (2) To apply statistical approaches where combining transcriptomics and proteomics increases power. (3) To help understand where and how proteomics and transcriptomics results differ from each other.

Results

The effort to visualize multiple data types together and to improve the quality of biological pathways has thus far resulted in the development of a new pathway visualization and editing tool named PathVisio (http;// This tool is designed to be a user-friendly viewer for high-throughput data as well as an editor for biological pathways. To incorporate multiple data types it was necessary to extend GPML, and XML-based pathway storage format based on the GenMAPP pathway markup language.

Future Research

It is currently possible to visualize multiple data types together on the computer. This approach has been used to examine previously published data by Washburn et al. As more experimental data from the nutritigenomics field becomes available, and pathway content is steadily improved, the focus will move from visualization and pathway content to applying these techniques to aid the analysis of experimental data.

References
Washburn MP et al., Proc Natl Acad Sci U S A 2003 Mar 18; 100(6) 3107-12.