Genome-wide association study (GWAS) results for Type 2 diabetes suggest a much larger footprint for islet cell dysfunction in T2D than previously thought. While the "insulin resistance" paradigm still works, we've had to adapt our model to include the more disordered insulin secretion indicated by these results. This is why unbiased and hypothesis-free research methods like GWAS are so powerful -- they aren't dependent on our preconceived notions of how things "should" be. A nice review reference: Herder et al. Eur J Clin Invest. 2011;41(6):679-92.

Totally agree. This is actual news for nerds who are interested in how to effectively manage, and be managed, by other nerds. Or, we could go back to arguing whether Autism is a fake diagnosis, based on, you know, our skill with Java. I kid, I kid. Sort of.

They didn't predict anything. They retrospectively reviewed scans and determined a "signature" that correlated with the outcome studied. Without an independent validation cohort, this is interesting but far short of definitive. There are concerns about overfitting with such an analysis technique.

I really think we should come up with a team to figure out how in-home cold fusion reactors can be integrated into the existing power grid. This is a pressing issue, and I know if we work together we can achieve seamless integration. How's 6pm on Thursday sound?

The term "personalized medicine" is a buzzword. We've been targeting specific environmental things in specific people for a long time now. It turns out it is hard to get people to get in shape, control their blood pressure, and quit their bad habits. Genetics is as personal as it gets, so that has become the new holy grail. Genetics offers the ability to identify new risk factors and improve understanding of the underlying disease. Many of the identified common and even rare variants in disease don't lead to therapies, but they do tell us about the pathobiology, which may in turn lead to new discoveries and/or therapies. We would love to study gene-environment interactions, but there are major power issues when trying to do studies like this in sporadic, complex disease like heart disease, hypertension, stroke, and the like. We're still trying to develop large enough sample databases and robust analytic techniques that allow us to study the interplay between millions of genetic variants and the often difficult to quantify or report environmental exposures that may act in concert in additive or even multiplicative ways.

The use of terms for sequence data and expression data are not interchangeable. The U133 microarray is for RNA, yes RNA, expression data. RNA microarrays quantify the fold change difference in expression between different subjects. DNA microarrays identify polymorphisms or repeats or the like. While arrays like the U133 rely on sequence level data to create the array, this is not the same as saying that sequence-level data is contaminated. Bottom line, the fact that this is not the cover article for Nature|Genetics this month tells you a lot of the story. Unless you are some sort of conspiracy theorist, or want to get swept up in the usual slashdot "sky is falling" imperative.

The wars of the future will not be fought on the battlefield or at sea. They will be fought in space, or possibly on top of a very tall mountain. In either case, most of the actual fighting will be done by small robots. And as you go forth today remember always your duty is clear: To build and maintain those robots.

So we can't stop our poor dumb natural immune systems from attacking our own bodies, and we're just a few short years away from telling it what to do?

If you just pick 24 random SNPs, they may not be particularly informative for your population (i.e. they may be monomorphic or have very low minor allele frequencies that don't help you discriminate individuals). So you want to pick markers that are bi or even tri-allelic with high MAFs for your population, to make sure they vary enough from person to person to tell them apart.