WP 14. Technologies - molecular bioinformatics and biostatistics

This area of work will be led by Partner 3 (Berlin-Charité) with input from all partners performing genome and transcriptome analyses, especially from Partner 1 (KI), Partner 2 (Leiden), Partner 16 (Birmingham), Partner 18 (Evry-Paris), Partner 21 (EULAR, associated network partners) and Partner 26 (oligene GmbH, Berlin).

Objective:

To develop, standardise and provide strategies for gene expression analysis of defined purified cell populations and complex mixtures of cells. To develop algorithms for comparison of different array platforms. To identify candidate genes for molecular diagnosis, classification, and therapy management of autoimmune arthritis and myositis on the basis of array analysis.

Workplan:

The first activity will be the definition of standards for the different array platforms and genetic analyses. This includes the establishment of standardized protocols for sample collection, processing and array hybridisation (in collaboration with WP12) as well as a standardized bioinformatics pipeline for data generation, processing and analysis. These protocols will be defined as a consensus between the different partners. Quality control evaluation will be guaranteed by the exchange of samples over the different microarray platforms.

After 6 months for setup of standards and comparability tests, profiles of defined cell types will be successively analyzed and compared in collaboration with WP12. In collaboration with WP3, 4, 6 and 7, analysis of gene expression data from early arthritis, established disease and therapeutic response will be started.

Furthermore, representative aliquots of tissue samples will be characterised in WP 4 prior to selection for expression analysis. For interpretation of array data of whole tissue samples (WP4), a new algorithm, FPCA, will be applied (Partner 3 and 26) to distinguish between genes associated with differential cellular composition versus candidates associated with molecular mechanisms of the disease.

To correlate molecular phenotypes with clinical parameters, a database for clinical assessment will be established by partner 26 which allows immediate linkage to molecular data.

To generate customized arrays, different selections of relevant genes and genes applicable for normalization procedures will be tested on a virtual basis.

Expected achievements at 18 months:

•  Generation of standards for analysis of DNA-microarrays in all three platforms.

•  Generation of algorithms to compare data between different platforms.

•  Transcriptomes and differences between transcriptomes of different cell populations including monocytes, T- and B-lymphocytes, macrophages, synoviocytes as well as whole tissue and blood.

•  An algorithm for diagnostic application in tissue samples to focus on genes associated with molecular mechanisms of disease and not with variability of cellular composition.

•  First candidate marker genes for molecular diagnosis and classification of autoimmune arthritis and myositis as well as for therapeutic stratification of arthritis.

•  Candidate genes selected for the production of customised cDNA arrays including monocyte/DC- oriented and T-cell oriented arrays.