Genome-wide association studies, which are the source for most of the data 23andMe uses in Gene Journal (now called Health and Traits) entries, are based on the “common disease-common allele model” — the idea that many illnesses, even relatively rare ones, can be caused by combinations of genetic variations that are individually quite common in the population.A study published online yesterday in Science Express suggests that for schizophrenia at least, and perhaps other mental disorders, that approach might not be the way to go.Schizophrenia is a debilitating psychiatric disorder that affects approximately one percent of the population. People with schizophrenia suffer from hallucinations, delusions, and disorganized thinking. The illness greatly impacts social and occupational functioning and has enormous public health costs.The authors of the paper propose that a genetic predisposition to schizophrenia is caused by structural variations in the genome such as deletions, duplications, and re-arrangements of genetic material instead of variation at SNPs. Furthermore, they think that these structural variations might be different for different patients, meaning that it would be difficult to ever find DNA markers that are predictive for the disease.The researchers used new technologies to look for structural variations in 150 people with schizophrenia or schizoaffective disorder and 268 healthy controls with no history of neurological or psychiatric illness. They found that individuals with schizophrenia were much more likely than controls to have structural variations that affected genes (as opposed to non-coding parts of the genome). The association was strongest in people who developed symptoms while 18 years old or younger.Virtually every structural variant the researchers detected in the individuals with schizophrenia was unique, though sometimes patients had differing mutations in the same genes. Genes involved in brain development were the most affected.The results of this study don’t prove that any one gene is associated with schizophrenia, but they do suggest that researchers who want to understand the genetics of this illness, and maybe other complex psychiatric disorders, should perhaps focus their efforts on structural variations instead of SNPs.That doesn’t mean SNPs aren’t useful. Almost every day we’re learning something new about how subtle single-letter DNA variations between people may affect their health. But some questions won’t be answered until scientists have a better understanding of many other types of genetic variation as well — not to mention the contribution of other factors such as diet, personal habits and environmental exposures.