A Dense (1 Million SNP) Genome-Wide Association Study in Pima Indians

我们使用Affymetrix 10万SNP芯片和100万SNP芯片技术完成了受试者的基因分型。第一阶段旨在通过对300名早发性2型糖尿病受试者（发病年龄<25岁）和329名非糖尿病对照（年龄0 ~ 45岁）以及271名被选受试者的糖尿病和非糖尿病兄弟姐妹进行基因分型，来检测与年轻发病2型糖尿病的潜在关联。使用病例/对照分析（N= 629）和家族内分析（来自169个兄弟姐妹的482个兄弟姐妹）计算与糖尿病的关联，并优先考虑与组合关联最强的snp。GWA的第2阶段旨在检测与糖尿病前期特征（体脂百分比，高胰岛素正糖钳技术测量的胰岛素作用，静脉注射葡萄糖的急性胰岛素反应）的关联。对600名非糖尿病受试者进行了这些糖尿病预测因子的代谢表型分析。在第一阶段和第二阶段的所有样本中都可以测量BMI。在3501个全遗传皮马印第安人群体样本中，选择与糖尿病和/或糖尿病前期特征（包括BMI）有最佳关联的snp进行额外的基因分型。这种基因分型利用了一种新的高通量技术（Bead Express）。我们最近完成了3501个皮马印第安人全基因组关联分析中最佳信号的基因分型，目前正在3784个混合血统印第安人的第二个基于人群的样本中复制该样本的最佳snp。迄今为止，在合并的糖尿病样本中，与DNER中的snp （P= 1 x 10-8）和KCNQ1中的snp （P= 5 x 10-9）的相关性最强。在合并样本中，BMI与BRD2、heat5b、UBE2E、GSTA5、NOVA1、FTO、MAP2K3和CYB5A的snp相关性最强（P值均在10-6和10-7之间）。目前正在进行研究，以确定每种关联背后的偶然变异。

We completed genotyping subjects with both the Affymetrix 100,000 SNP chip and the 1 million SNP chip technologies. Phase 1 was designed to detect potential associations with young-onset type 2 diabetes, by genotyping 300 early-onset type 2 diabetes subjects (onset age<25 yrs) and 329 non-diabetic controls (age >45 yrs), and 271 additional subjects who were diabetic and non-diabetic siblings of the selected subjects. Associations with diabetes were calculated using both a case/control analysis (N= 629) and a within-family analysis (482 siblings from 169 sibships), and SNPs that had the strongest association for the combined associations were prioritized. Phase 2 of the GWA was designed to detect associations with pre-diabetic traits (% body fat, insulin action as measured by the hyperinsulinemic euglycemic clamp technique, and the acute insulin response to an intravenous bolus of glucose). Six hundred non-diabetic subjects who had been metabolically phenotyped for these predictors of diabetes were genotyped. Measures of BMI were available on all samples from Phase 1 and Phase 2. SNPs that provided the best associations for diabetes and/or a pre-diabetic trait (including BMI) were selected for additional genotyping in a population-based sample of 3501 full-heritage Pima Indians. This genotyping utilized a new high throughput technology (Bead Express). We recently completed genotyping the best signals from our genome-wide association analysis in a sample of 3501 full hertiage Pima Indians and are currently replicating the best SNPs from this sample in a second population-based sample of 3784 mixed heritage Native Americans. To date the strongest assocations in the combined samples for diabetes are with SNPs in DNER (P= 1 x 10-8) and with KCNQ1 (P= 5 x 10-9). The strongest associations in the combined samples for BMI are with SNPs in BRD2, HEATR5B, UBE2E, GSTA5, NOVA1, FTO, MAP2K3, and CYB5A (all P between 10-6 and 10-7). Studies are ongoing to identify the casual variant that underlies each of these associations.