Microarray data quality is an issue that has been covered extensively in The Scientist (see, for instance, here and here). The basic issue is this: how reliable are the sometimes-subtle changes in gene expression levels these experiments yield, and how reproducible are they.
The External RNA Controls Consortium, an ad-hoc group with approximately 70 members from private, public, and academic organizations formed in 2003, is one of several groups working to address these questions. The Consortium?s goal is to develop a reference set of external RNA control (or ?spike?) transcripts, informatics tools, and procedures that can be used to assess technical performance in gene expression assays like microarrays and quantitative, real-time RT-PCR.
If that sounds boring, it shouldn't: the upshot is greater confidence in gene expression data, not to mention the potential for accelerated regulatory approval of microarray diagnostics and accreditation of testing labs.
Recently the consortium released...
Here?s how the Consortium described the testing plan in a progress report published in the October Nature Methods:
?Generally the test plan consists of testing the controls across multiple platforms at several member sites?. All sites will run agreed-upon protocols and analysis methods. The type of information collected during the testing includes: crosshybridization, quantitative performance titration experiments, performance in complex samples and independent confirmation of performance.?
Next up is an open ERCC Test Plan meeting, Oct. 4 and 5 at the Lister Hill Auditorium at the NIH, and testing is scheduled to begin at the end of 2005, with analysis extending into the new year. The report concludes, "The ERCC welcomes the participation of anyone in the community with an interest in helping to drive these efforts forward."
I encourage anyone with the time and interest to contact the ERCC at email@example.com. For years, we've heard about the potential of microarrays. With the ERCC's help, they might just make good on that promise.