Enter Note Done

Session


Keywords: Genome-wide association; Obesity; Genotype-phenotype correlations; SNP analysis/discovery

Authors:
R.W. Read 1,2; K.A. Schlauch 1,2; E.T. Cirulli 3; N.L. Washington 3; S. White 3; J. Lu 3; A. Slonim 1,2; J.J. Grzymski 1,2

Affiliations:
1) Institute for Health Innovation, Desert Research Institute, Reno, NV, USA; 2) Institute for Health Innovation, Renown Health, Reno, NV, USA; 3) Helix, San Carlos, CA, USA


The aggregation of Electronic Health Records (EHRs) and personalized genetics has led to novel discoveries relevant to population health. Here we use the Illumina Human OmniExpress-24 BeadChip and Helix Exome+ sequencing in combination with EHRs from Renown Health in Reno, NV to identify variants associated with BMI and severe class II obesity. Three genome-wide association studies (GWASs) of 500,000 variants on the Illumina platform of 6,645 Healthy Nevada Project participants identified published and novel variants that associated with BMI and obesity. After standard quality control measures, with age and gender as covariates, the first GWAS identified 20 variants associated with BMI (p<1x10-5). When adding Type II Diabetes (DM2) as a comorbidity in the second GWAS, 27 significant variants were identified, with several in common with the BMI-only GWAS. We found 12 variants within the FTO gene to be significantly associated with the quantitative BMI trait, an association which has been illustrated in many prior studies. This association was further validated by the Helix Exome+ sequencing of 20,000 additional Healthy Nevada Project (HNP) participants using the same statistical model, in which seven top associations (mean p-value 7x10-13) are variants in FTO. In the original GWAS cohort, TDH showed strong associations with BMI (five variants), likely through its secondary association to DM2. TDH is a pseudogene, which may be transcribed to an siRNA that regulates gene expression, and is not yet known to have a link to BMI. A third complementary case-control GWAS was performed to examine links with extreme obesity: this identified variants already linked to BMI, with more variants identified in NEGR1 than in the quantitative BMI GWAS. NEGR1 likely plays a role in body weight regulation, as identified by knockout mice in prior studies. A novel variant in PFKFB3, a gene not yet directly linked to BMI, was strongly associated with obesity in this case-control study and may be protective of developing obesity. Two phenome-wide association studies (PheWASs) examined phenotypic links to BMI and clinical associations to our obesity-associated variants. The first PheWAS identified strong associations between BMI and obesity to DM2, hyperlipidemia, hypertension, sleep apnea, GERD, and asthma. The second showed associations between SNPs in TDH, NEGR1, and FAM167A-AS1 to DM2, and SNPs in CABP5 to IBS. These studies highlight prior and novel links of BMI and obesity.