Remember SARS? It came, there was a wealth of "end of the world is nigh" TV documentaries, some guys in Canada sequenced it over a weekend...and then it was yesterdays news.
The latest BMC Genomics contains an article (link) which claims that the clinical severity of SARS correlates with the expression of around 52 signature gene transcripts from whole blood samples.
I am no expert in the array analysis methods that they've used or medicine in general, but it looks interesting.
Comments
Why the shock?
Just curious. Is it your belief that microarrays generally aren't useful?
Belief
Who cares what I believe? They're all the rage, so they must be good :)
Since you asked though - my personal feeling is that microarrays are a great idea in principle but are rather let down by a lot of the practice. If your northern or southern blot is afflicted by high background or poor visualisation, you spend some time optimising the hybridisation conditions. Conversely, when people have the same problems with microarrays (which to me are just northern blots on steroids), they instantly leap to a wide assortment of complex statistical methods to massage their data. I'd have more faith in microarray results if as much time was put into improving experimental technique and design as goes into normalisation.
The Australian academic community is also plagued by people who confuse hype with substance and tend to latch onto techniques, then equate them with the whole field - "microarrays = bioinformatics", "proteomics = bioinformatics" and so on. I think it's because they can't conceive of what standalone computational analysis entails and need to tie it to what they understand, i.e. molecular biology.