and Custom Protocols for Microarray Analysis
CII Project Staff
Bérengère Bouzou, Georgios Asteris, Yong Gao and Tim Clark
Collaborators
Michael Reich and Ted Liefeld (Broad Institute for Biomedical Research)
Figure 1 - from Reich et al. (2004) GenePattern Tutorial and Reference
Description
CII staff are working to implement a Gene Pattern Server along with a set of custom-designed protocols to support self-service analysis of microarrys for typical experiments run by biologists at MIND. We will provide a simple interface for biologists and clinicians to implement standardized, robust analytical pipelines for their microarray experiments. The “protocols” are sequentially-executed analytical methods formed into a “pipeline” construct: output from each step is input to the next step (Figure 1). These protocols are being developed by CII informaticians specifically for use at MIND with the GenePattern analysis framework developed by Reich et al. at the Broad Institute.
Gene Pattern provides biologists, researchers and clinicians with little or modest computational backgrounds with flexibility to run well-constructed multi-step statistical analyses on microarray expression data, and look at results without writing programs themselves or finding extensive help from a computational colleague.
In general all the statistical routines from which we construct pipeline protocols are standard package routines in Bioconductor/R or MatLab. There will be occasional exceptions for special in-house-developed code for special putposes.
The Gene Pattern server and protocol library is being implemented on the MIND-F1 computational server (see Infrastructure project). We expect it to be released for beta users towards the end of summer, 2005.
Participate as a Design Partner or a Beta User
Design Partners collaborate with us to design the analysis protocols, and supply experimental data for the analysis while working with us to optimize the design for their research.
Beta Users are early-release volunteer users who gain early access to the protocols, in exchange for feedback. They work with us to optimize the protocols for a wider audience.
To participate in this project, contact Bérengère Bouzou or Tim Clark.
Funding
This project is funded by grants from the Udall Centers of Excellence in Parkinson’s Disease Research and a private charitable foundation.