Genome-wide association studies (GWAs) have received a lot of media attention in the last several months as various research groups have released over a half-dozen such studies, all focused on some of the most widespread Western diseases, including heart disease, type II diabetes, and breast cancer. (See here, here, and here for some examples.) These studies have the potential to substantially change how we understand, diagnose and treat these diseases, and they possibly signal the near-arrival of at least some of the long-promised health benefits of the Human Genome Project, although new cures are probably still many years in the future.
Genome-wide association studies have become feasible with the availability of the human genome sequence and several associated technologies that allow researchers to rapidly and extensively genotype large numbers of people (I have previously explained how these studies work here). Traditional studies have compared phenotypes of patient populations with healthy subjects, in an attempt to correlate certain lifestyle risk factors with disease, such as studies linking diet with heart disease. These traditional studies are important, but until recently we could only phenotype patients, not genotype them. Now, with new technologies available, we can compare sick and healthy subjects at thousands of places in their genomes, and identify genetic variants that are possibly linked with a disease.
These recently reported GWAs are just a first pass, and their results will take some time to pursue in more depth. But with each study we are finding a handful of genes (or more accurately, variants of genes, called alleles) that may be involved in disease. These alleles could very quickly become useful for identifying people who are at high risk for disease.
As far as new treatments go however, I wouldn't expect to see many in the near future. The problem is this: GWAs give us a list of genetic variants linked to a disease, yet we have no idea what most of these variants do. We still have to understand how those variants affect specific proteins in the cell (i.e., molecular biology), and how those variants lead physiologically to disease (i.e., pathogenesis). Finally, we have to understand how multiple genetic variants act in combination to produce complex diseases like diabetes. Unlike Mendelian diseases, like cystic fibrosis which can be linked to a defect in a single gene, complex diseases like diabetes involve multiple genes interacting in a complex way with environmental factors. We hardly know how to study such complex interactions, much less cure them. This is an extremely difficult scientific problem.
Does this mean that these cures will never come? That the Human Genome Project was a waste of money and effort? No way - genome sequencing projects have already been a huge boon to basic science research, and they are also the obvious way forward in biomedical research. Molecular biology and genomics have not yet produced health benefits on par with the germ theory of disease, vaccinations, and randomized double-blind studies, but some day they will.