By Julia Karow
This article was originally published Nov. 8.
WASHINGTON, DC – In an attempt to determine the role of low-frequency coding variants in common metabolic diseases like type-2 diabetes and obesity, researchers at BGI in Shenzhen and in Denmark have sequenced the exomes of 1,000 Danish disease samples and 1,000 controls, as well as validated interesting variants in another 10,000 disease samples and 7,000 controls.
At the American Society for Human Genetics meeting last week, Oluf Pedersen, head of LUCAMP, the consortium conducting the study, and Yingrui Li, director of the bioinformatics center at BGI Shenzhen, presented an outline of the project goals and methods. The scientists are currently analyzing the data and hope to obtain first results within a few months.
LUCAMP (a shortening of the Lundbeck Foundation Center for Applied Medical Genomics in Personalized Disease Prediction, Prevention and Care) involves BGI as well as groups at the Hagedorn Research Institute, the Steno Diabetes Center, the University of Aarhus, and the University of Copenhagen. The project started in 2007 and receives $12 million in funding from the Lundbeck Foundation as well as several million dollars in additional funding from BGI.
One goal of the project is to identify variants in coding regions of the genome — with a focus on rare variants with an allele frequency between 1 and 5 percent — that confer an increased risk for metabolic disease, such as type-2 diabetes or obesity. For a pilot project that was published last month, the researchers initially sequenced the exomes of 200 healthy Danish individuals and found that the low-frequency variants harbored an excess of deleterious mutations (IS 10/5/2010).
For the full project, they have now sequenced the exomes of 1,000 Danish patients, each suffering from visceral obesity, type-2 diabetes, and hypertension, as well as the exomes of 1,000 glucose-tolerant and lean so-called "super-controls," also from Denmark.
The researchers used NimbleGen capture arrays, which target about 18,500 genes, to enrich protein-coding DNA, which they sequenced at about 10-fold coverage on the Illumina Genome Analyzer. In total, they called approximately 150,000 SNPs.
In a second stage, they validated 20,000 of those SNPs by genotyping them in 5,000 Danish patients with type-2 diabetes and 5,000 glucose-tolerant controls, as well as in 5,000 obese Danish individuals and 5,000 lean controls. For this, they used an Illumina iSelect custom-designed genotyping array, in which they included the 1,000 highest-ranked candidate SNPs, 200 SNPs known to be associated with metabolic disease, as well as all non-synonymous SNPs.
While BGI sequenced and genotyped the samples, the data analysis, which is currently ongoing, is performed in Denmark. The analysis will include comparisons between cases and controls of genes in which SNPs were found at multiple locations, as well as pathway studies.
The researchers may also attempt to replicate the results in another cohort, probably from Sweden. Replication studies in non-Danish ethnic groups might be difficult, Pedersen explained, because the low-frequency variants only appeared recently in evolution.
Ultimately, he said, it will be important to also investigate variants with a frequency of less than 1 percent. This could be done, for example, through deep exome sequencing of families with several members affected by the disease.
While there are "no conclusive results" yet, and the study might have missed variants in regions that were not captured, it will hopefully shed light on the role of rare variants in common metabolic disease, and Pedersen said he hopes that other groups will follow suit in exploring them.
Have topics you'd like to see covered in In Sequence? Contact the editor at jkarow [at] genomeweb [.] com